The first half of the identity chapter, all of it descending from one distance-formula computation. Covers the even-odd identities and their proof by reflection across the x-axis, the difference identity for cosine proved from equal chords, the cofunction identities that fall out of it and explain the prefix co, the sum and difference identities for sine derived through the cofunction identities, and the tangent formula obtained by dividing. Includes finding exact values of non-special angles by decomposing them into special ones.
Subject: Trigonometry · 70 slides · symbolic lesson
Open the interactive version of this deck
Title
Trigonometry · Section 10.4
One chord, one distance formula, and everything else follows
Objectives
Section 10.3 gave you the Pythagorean, quotient and reciprocal identities: all of them relate the six functions at one angle. This lesson relates them at different angles, which is what lets you reach angles the unit circle never listed.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 770-776 — Theorem 10.12 through Theorem 10.16
Everything you have met so far relates the six functions at a single angle. The Pythagorean identity ties the cosine and the sine of the same angle together; the quotient and reciprocal identities do the same. None of them lets you say anything about a new angle.
This lesson changes that. Its five theorems relate values at different angles: at an angle and its negative, at an angle and its complement, and at a sum or difference of two angles. That last one is the big one, because it is what lets you compute exact values at angles the unit circle never listed, such as fifteen degrees or nineteen twelfths of pi.
There is far less to memorise here than there appears to be. One theorem is proved from a picture, and the other four are squeezed out of it in a line or two each. If you follow the chain of derivations once, you will be able to rebuild any formula you forget, which is much safer than trusting a sign you half remember.
Warm-up
Discussion prompt
Before anything new: is the cosine of a sum the sum of the cosines? Test it with the two angles below, where you know every value involved, and write down both numbers.
Write both numbers down before revealing. Committing to an answer, even a wrong one, is what makes the correction stick, and this particular wrong answer is worth having on paper.
The temptation is real because so much of algebra does distribute. But cosine is a function, not a coefficient. Feeding it a sum does not split it any more than squaring a sum splits, or taking a logarithm of a sum splits. The two numbers here are zero and about 1.366, which is not close in any sense.
So there has to be a formula, and it will not look like distribution. It turns out to involve four products of values at the individual angles, and the rest of this lesson is about where it comes from and how to keep its signs straight.
Section
Part 1
Concept
Rotate by a positive angle and you go counter-clockwise; rotate by its negative and you go the same amount clockwise. The two terminal points are mirror images in the horizontal axis.
Figure (svg): A unit circle with a ray at fifty degrees reaching a point P above the horizontal axis and a ray at minus fifty degrees reaching a point Q directly below it; a dashed vertical segment joins P to Q, and small arcs mark the two angles.
\[ \text{reflection in the } x\text{-axis: } (x, y) \mapsto (x, -y) \]
A reflection in the horizontal axis keeps the first coordinate and negates the second. The first coordinate is the cosine and the second is the sine, so that single sentence is the theorem.
\[ \cos(-\theta) = \cos(\theta), \qquad \sin(-\theta) = -\sin(\theta) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 770-771 — the proof of Theorem 10.12
Look at the picture and be clear about what a negative angle means. It is not a negative number that gets measured out; it is a rotation in the other direction. Sweeping fifty degrees counter-clockwise and fifty degrees clockwise lands you at two points that are mirror images in the horizontal axis.
Now recall the definition from Section 10.2: the cosine of the angle is the first coordinate of that terminal point and the sine is the second. A reflection in the horizontal axis leaves the first coordinate exactly where it was and negates the second. That is the entire content of the theorem, and it is worth being able to say it in that one sentence.
The dashed segment in the figure is the reflection. Notice that the two points sit directly above and below each other, at the same horizontal position. If you can see that, you will never again have to remember which of cosine and sine keeps its sign.
Concept
The remaining four functions are built from the cosine and the sine, so their behaviour is decided by counting how many sine factors each one carries.
\[ \sec(-\theta) = \frac{1}{\cos(-\theta)} = \frac{1}{\cos\theta} = \sec\theta \]
\[ \csc(-\theta) = \frac{1}{\sin(-\theta)} = \frac{1}{-\sin\theta} = -\csc\theta \]
\[ \tan(-\theta) = \frac{\sin(-\theta)}{\cos(-\theta)} = \frac{-\sin\theta}{\cos\theta} = -\tan\theta \]
\[ \cot(-\theta) = \frac{\cos(-\theta)}{\sin(-\theta)} = \frac{\cos\theta}{-\sin\theta} = -\cot\theta \]
even and odd — A function is even when negating the input leaves the output alone, and odd when negating the input negates the output. Cosine and secant are even; sine, cosecant, tangent and cotangent are odd.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 770 — Theorem 10.12
The book proves only the cosine and sine cases and leaves the other four as exercises, because each is two lines. Work through the four lines on this slide rather than reading them, and notice that they all have the same shape.
Here is the counting rule that makes the answer memorable. Write each function in terms of cosine and sine, then count the sine factors. Secant has none, so nothing changes sign and it is even. Cosecant, tangent and cotangent each have exactly one, so each changes sign and all three are odd.
The result is lopsided: two even functions and four odd ones. That is not an accident of naming. Only cosine and secant read the horizontal coordinate, and the horizontal coordinate is the one the reflection leaves alone.
Notation
Annotate
On: \( \cos(-\theta) = \cos\theta, \quad \sec(-\theta) = \sec\theta, \quad \sin(-\theta) = -\sin\theta \)
Step through the annotations and connect each one to the picture on the earlier slide. The most important thing to fix is where the minus sign lives. It is inside the bracket, attached to the angle, which means the claim is about a rotation direction, not about a negative output.
The final annotation is the one that quietly matters. Every theorem in this section says for all applicable angles, which means wherever both sides are defined. The tangent version is silent at a quarter turn, because neither the tangent of that angle nor the tangent of its negative exists. Saying an identity holds for all angles when it does not is a cheap way to lose a mark.
One more habit worth starting here: whenever a theorem gives you six statements, look for the two that carry all the information. Here it is the cosine one and the sine one, and the rest are bookkeeping.
Picture it
Figure (svg): Two graphs side by side over minus two pi to two pi. Left: the cosine curve with two dots at plus and minus 1.3 at the same height. Right: the sine curve with two dots at plus and minus 1.3 at opposite heights.
The same fact, seen as a symmetry of the whole graph rather than one point. The cosine graph is unchanged by folding the page along the vertical axis; the sine graph is unchanged by turning the page half a revolution about the origin.
This is the same fact, lifted from a single point to the whole graph. On the left, the two dots sit at the same height because the cosine of an angle and the cosine of its negative are equal, so the graph is unchanged when you fold the page along the vertical axis.
On the right the two dots sit at opposite heights. The sine graph is not symmetric about the vertical axis at all, but it is unchanged if you rotate the page half a turn about the origin. That is what odd means as a picture.
Being able to switch between the two views costs nothing and pays often. When a problem gives you a graph, the symmetry is the fastest way to see which identity applies; when a problem gives you an angle, the unit circle picture is faster. They are the same statement.
Intuition
Most of this section's exercises say verify the identity, and that is a different instruction from solve the equation. The first verification is one slide away, so it is worth being precise now.
\[ \text{an equation: } \sin\theta = \tfrac12 \qquad \text{true for some angles} \]
\[ \text{an identity: } \sin(-\theta) = -\sin\theta \qquad \text{true for every applicable angle} \]
Solving an equation means finding the angles that make it true. Verifying an identity means showing that there are none to find, because it already holds everywhere both sides are defined.
The method follows from that. Work on one side only, usually the more complicated one, and transform it into the other using identities you have already proved. Do not move terms across the equals sign, and do not multiply both sides by anything, because either move quietly assumes the very equality you are trying to establish.
This is a distinction that costs marks when it is blurred, so fix it before the first verification. An equation is a question: which angles make this true? An identity is a claim: this is true at every angle where both sides make sense. Verifying one means proving the claim, not hunting for solutions.
The method is forced by the logic. You may take one side and transform it, step by step, using facts already established, until it becomes the other side. Every line you write is then equal to the one above it, and the chain proves the claim.
What you may not do is treat the identity as an equation you are allowed to operate on. Moving a term across the equals sign, or multiplying both sides by something, assumes the two sides are already equal, which is exactly what you were asked to show. It is circular, and it is the most common way a verification loses full marks even when the final line is correct.
Worked example
The two sides look different only because a factor of minus one is hidden inside the argument on the left.
\[ \text{Verify: } \cos\left(-\frac{\pi}{4} - 5t\right) = \cos\left(5t + \frac{\pi}{4}\right) \]
Factor minus one out of the argument on the left
Why: Both terms inside the bracket carry a minus sign.
\[ -\frac{\pi}{4} - 5t = -\left(5t + \frac{\pi}{4}\right) \]
Apply the even identity for cosine
Why: Negating the input of a cosine changes nothing.
\[ \cos\left(-\left(5t+\frac{\pi}{4}\right)\right) = \cos\left(5t+\frac{\pi}{4}\right) \]
Check at a value
Why: At t equal to 0.3 the argument on the left is minus 2.285398 and on the right it is 2.285398; both cosines come out as minus 0.655317, so the two sides agree.
\[ \cos(-2.285398) = -0.655317 = \cos(2.285398) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 782 — Exercise 2
The only skill here is spotting that a factor of minus one can be pulled out of the argument. The left side has two terms inside the bracket and both of them are negative, so the whole argument is minus one times something. Once you see that, the even identity finishes it in one move.
The rest of the section is full of this. When an expression looks unlike anything in the theorem list, look first for a hidden minus sign inside an argument, clear it, and look again. That is why the pattern slide at the end of the lesson puts this step first.
The numerical check is worth doing even on a verification. Pick an ugly value for the variable, evaluate both sides, and compare. If the two sides differ, you have found an error without hunting for it; if they agree at a randomly chosen value, your algebra is almost certainly right.
Worked example
The book says the real use of these identities is not evaluating but simplifying. Here three minus signs are hiding in one product.
\[ \text{Simplify: } \sin(-\theta)\cos(-\theta)\cot(-\theta) \]
Apply the even-odd identity to each factor
Why: Sine and cotangent are odd; cosine is even.
\[ = \left(-\sin\theta\right)\left(\cos\theta\right)\left(-\cot\theta\right) \]
Cancel the two minus signs
Why: A negative times a negative is positive.
\[ = \sin\theta\,\cos\theta\,\cot\theta \]
Rewrite the cotangent as a quotient
Why: Cotangent is cosine over sine.
\[ = \sin\theta\,\cos\theta\cdot\frac{\cos\theta}{\sin\theta} \]
Cancel the sine
Why: Legitimate wherever the sine is nonzero, which is exactly where the cotangent was defined.
\[ = \cos^2\theta \]
Check at a value
Why: At theta equal to 0.8 radians the original product is 0.485400 and the cosine squared of 0.8 is 0.485400.
\[ \sin(-0.8)\cos(-0.8)\cot(-0.8) = 0.485400 = \cos^{2}(0.8) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 771 — the remark on simplifying, before Theorem 10.13
The book is explicit that these identities are not really for computing values. Their use is simplification, and this example shows the shape that use takes: a product with several negated arguments, which become several minus signs, which then cancel in pairs.
Count the minus signs before you do anything. There are three negated arguments, but only two of them produce a minus sign, because cosine is even. Two minus signs cancel, so the answer will be positive, and knowing that in advance protects you from the most common slip.
The last two steps are ordinary algebra with the quotient identity, and the cancellation of the sine is legitimate precisely where the original expression made sense: the cotangent already required a nonzero sine. Whenever you cancel something, it is worth saying out loud why the thing you cancelled could not have been zero.
Fill the middle
Theorem 10.12 leaves the last four identities as exercises. Rebuild the cotangent one from the quotient identity.
Fill in the blanks
\cot(-\theta) = \frac\cos\theta-\sin\theta = \frac-\cot\theta}___} = ___
Why: The cosine on top is even, so its minus sign disappears. The sine underneath is odd, so its minus sign survives. One minus sign in a quotient makes the whole quotient negative, so the cotangent is odd. The same two lines with the reciprocal identity prove that the cosecant is odd and the secant is even.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 770 — the four identities left as exercises
Fill both blanks before checking. This is the argument the book leaves as an exercise, and it takes two lines, so there is no reason to look it up rather than rebuild it.
The top and the bottom behave differently, and that is the whole point. The cosine on top is even, so its minus sign simply disappears. The sine underneath is odd, so its minus sign survives, and a single minus sign in the bottom of a fraction makes the whole fraction negative.
Run the same two lines with the reciprocal identity instead and you get the secant and cosecant cases. That is four of the six identities proved in about six lines, which is the sort of return that makes it worth learning derivations rather than lists.
Sorting
Sort into buckets
Sort the six circular functions. Ask which coordinate each one reads, then ask what the reflection does to that coordinate.
Sort all six before revealing. Do not sort from memory; sort by asking, for each function, which coordinate it reads, and then what the reflection does to that coordinate.
Two go in the even bucket and four in the odd one. If you found yourself putting three in each, you were probably pattern-matching on the prefix co, which does not help here: cosine is even but cotangent and cosecant are both odd.
The one worth double-checking is the tangent. It reads both coordinates, so it is tempting to guess that the two effects cancel. They do not: the numerator changes sign and the denominator does not, so the quotient changes sign.
Trap
A line that turns up whenever a minus sign appears inside a bracket:
\[ \cos\left(-\frac{\pi}{6}\right) = -\frac{\sqrt3}{2} \]
Wrong. Cosine is even, so the minus sign does not survive at all.
A negative angle is a clockwise rotation, not a negative number being measured. Minus a sixth of a turn lands in quadrant four, where the cosine is positive. Only the sine changes sign.
\[ \cos\left(-\frac{\pi}{6}\right) = \frac{\sqrt3}{2} \]
\[ \sin\left(-\frac{\pi}{6}\right) = -\frac12 \]
This error comes from reading the minus sign as an instruction about the answer rather than about the rotation. It is the same confusion that makes people expect the square of a negative number to be negative.
Fix it with the picture instead of a rule. Minus a sixth of a turn is a small clockwise rotation, which lands in the fourth quadrant. In the fourth quadrant the first coordinate is positive and the second is negative, so the cosine is positive and the sine is negative. That reading also tells you the sizes, not just the signs.
A useful test for any identity of this kind: evaluate at an angle where you know both answers, and see which version survives. Minus a sixth of a turn works well, because all six functions are easy there.
Socratic
The book does not prove the even-odd identities for a general angle straight away. It first replaces the angle with a coterminal one between zero and a full turn.
Discussion prompt
Why is that replacement needed, and why does it not weaken the result?
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 770-771 — the coterminal reduction in the proof
Write an answer before revealing. The question is really about what a picture is allowed to prove, and it is a habit worth building early, because the same reduction shows up again in the chord proof two slides later.
The picture shows one sweep from the positive horizontal direction to a terminal side. An angle of, say, twenty-three revolutions plus a bit has the same terminal side but cannot be drawn as one sweep, so the figure does not literally depict it. The reduction replaces such an angle by the one between zero and a full turn that shares its terminal side.
The reason this costs nothing is the second line on the slide. Negating a coterminal angle gives a coterminal angle, because negating a whole number of revolutions still leaves a whole number of revolutions. So the reduced pair and the original pair have exactly the same cosines and sines, and an identity proved for one holds for the other.
Section
Part 2
Concept
Only one theorem in this section is proved from geometry. Everything else is squeezed out of it algebraically, so this is the slide to understand.
Figure (svg): Two unit circles. On the left, rays at 105 and 35 degrees with points P and Q joined by a thick chord. On the right, a ray at 70 degrees with point A joined by an equally long chord to the point B at one, zero.
Take two angles and look at the arc between them. Now slide that arc round so that it starts at the point one, zero. The straight-line distance between its two ends has not changed, because the rotation was rigid.
\[ |PQ| = |AB| \quad \text{where } P, Q, A, B \text{ lie on the unit circle} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 771-772 — the picture before Theorem 10.13
Spend a moment on this slide, because it is the only geometry in the lesson and everything else descends from it. The claim is as simple as it looks: if you take the arc between two points on a circle and rotate the whole arc rigidly, the straight-line distance between its endpoints does not change.
In the left panel that distance is between the point at alpha and the point at beta. In the right panel the same arc has been rotated so that one end sits at the point one, zero, and the other end is now at the angle alpha minus beta. Both chords subtend an angle of seventy degrees at the centre, so both have the same length.
What makes this useful is that the distance formula turns a geometric equality into an algebraic one. Two expressions in cosines and sines are forced to be equal, and when you simplify them with the Pythagorean identity, almost everything collapses and the difference formula falls out.
Worked example
One equality of lengths, squared, and simplified with the Pythagorean identity. This is the only place in the lesson where a picture does any work.
Write the square of the first chord with the distance formula
Why: The two ends are the points at alpha and at beta.
\[ |PQ|^2 = (\cos\alpha - \cos\beta)^2 + (\sin\alpha - \sin\beta)^2 \]
Expand both squares
Why: Each bracket gives a square, a cross term and a square.
\[ = \cos^2\alpha - 2\cos\alpha\cos\beta + \cos^2\beta \]
\[ \qquad + \sin^2\alpha - 2\sin\alpha\sin\beta + \sin^2\beta \]
Apply the Pythagorean identity twice
Why: The cosine squared plus the sine squared of each angle is one.
\[ |PQ|^2 = 2 - 2\cos\alpha\cos\beta - 2\sin\alpha\sin\beta \]
Write the square of the second chord the same way
Why: Its ends are the point at alpha minus beta and the point one, zero.
\[ |AB|^2 = \left(\cos(\alpha-\beta) - 1\right)^2 + \sin^2(\alpha-\beta) \]
Expand and apply the Pythagorean identity once
Why: Now at the single angle alpha minus beta.
\[ |AB|^2 = 2 - 2\cos(\alpha-\beta) \]
Set the two squares equal
Why: The chords are the same length, so their squares are too.
\[ 2 - 2\cos(\alpha-\beta) = 2 - 2\cos\alpha\cos\beta - 2\sin\alpha\sin\beta \]
Subtract two and divide by minus two
Why: Dividing by a negative number reverses nothing here, because this is an equation.
\[ \cos(\alpha-\beta) = \cos\alpha\cos\beta + \sin\alpha\sin\beta \]
Check at a pair of angles you know
Why: Take alpha a third of a turn and beta a sixth. The left side is the cosine of a sixth of a turn; the right side is root three over four twice.
\[ \cos\frac{\pi}{6} = \frac{\sqrt3}{2}, \qquad \frac12\cdot\frac{\sqrt3}{2} + \frac{\sqrt3}{2}\cdot\frac12 = \frac{\sqrt3}{2} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 771-772 — the proof of Theorem 10.13
Follow this derivation once with a pencil, because it is the only proof in the lesson you cannot get by shuffling other formulas. Every line is either the distance formula, the expansion of a square, or the Pythagorean identity.
Watch what the Pythagorean identity does. After expanding, the left side contains four squared terms, and they pair up into two copies of one, which is where both of the twos come from. The right side has only two squared terms, which give one copy of one, and that is why the two sides simplify to such similar shapes.
Once both sides read two minus twice something, the conclusion is immediate. Subtract the two, divide by minus two, and the identity is on the page. The book handles the case where beta is bigger than alpha separately, and closes it with the even identity for cosine, which is another place the first part of this lesson earns its keep.
Notation
Annotate
On: \( \cos(\alpha - \beta) = \cos\alpha\cos\beta + \sin\alpha\sin\beta \)
Two features of this formula carry all the marks, and the annotations name them. First, the terms are matched: cosine goes with cosine and sine with sine, and nothing is mixed. Second, the sign in the answer is the opposite of the sign in the bracket.
The sign reversal is the single most missed detail in this section, so it is worth attaching a reason to it rather than a mnemonic. In the chord derivation, the difference formula came out with a plus, and the sum formula was then obtained by negating beta, which flipped the sine term. The reversal is a consequence of the sine being odd.
The last annotation matters when you start applying this in Chapter 11. There is no restriction on the angles at all. The coterminal reduction in the proof means the formula holds for angles of any size and either sign, which is what lets you use it on an angle of nineteen twelfths of pi without a second thought.
Concept
No second picture is needed. A sum is a difference with a negated second angle, and the even-odd identities clear up the negation.
\[ \cos(\alpha + \beta) = \cos\left(\alpha - (-\beta)\right) \]
\[ = \cos\alpha\cos(-\beta) + \sin\alpha\sin(-\beta) \]
\[ = \cos\alpha\cos\beta - \sin\alpha\sin\beta \]
So the two formulas are one formula. The terms are matched in both, and the sign in the answer is always the opposite of the sign in the bracket.
Theorem 10.13 — For all angles: the cosine of a sum is the product of the cosines minus the product of the sines, and the cosine of a difference is the product of the cosines plus the product of the sines.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 772 — Theorem 10.13 and the derivation of the sum case
This is the first of four places in the lesson where a new formula costs one line instead of a new proof, and the pattern is always the same: rewrite the thing you want in terms of the thing you have.
A sum is a difference in disguise. Subtracting the negative of beta is the same as adding beta, so the difference formula applies directly, and then the even-odd identities clear the two negated angles. The cosine term survives untouched and the sine term picks up a minus sign, which is exactly the reversal.
It is worth noticing that this means you only ever need to remember one of the two cosine formulas. If you are confident about the difference version, the sum version is one substitution away, and the same trick gets you the sine and tangent difference formulas from their sum versions later.
Worked example
Fifteen degrees is not on the unit circle's list, but it is the gap between two angles that are.
\[ \text{Find the exact value of } \cos(15^\circ). \]
Write 15 degrees as a difference of special angles
Why: Forty-five and thirty are both on the list, and their difference is fifteen.
\[ 15^\circ = 45^\circ - 30^\circ \]
Apply the difference identity
Why: Matched terms, and the minus in the bracket becomes a plus.
\[ \cos(45^\circ - 30^\circ) = \cos 45^\circ\cos 30^\circ + \sin 45^\circ\sin 30^\circ \]
Substitute the four known values
Why: Root two over two, root three over two, root two over two, one half.
\[ = \left(\frac{\sqrt2}{2}\right)\left(\frac{\sqrt3}{2}\right) + \left(\frac{\sqrt2}{2}\right)\left(\frac12\right) \]
Multiply each product out
Why: Numerators with numerators, denominators with denominators.
\[ = \frac{\sqrt6}{4} + \frac{\sqrt2}{4} \]
Combine over the common denominator
Why: Both terms already sit over four.
\[ = \frac{\sqrt6 + \sqrt2}{4} \]
Figure (svg): The cosine curve plotted from zero to ninety degrees, with dots at zero, thirty, forty-five, sixty and ninety degrees and a larger highlighted dot at fifteen degrees just below the value one.
Check the decimal
Why: Root six is 2.449490 and root two is 1.414214, so the answer is 3.863703 over 4, which is 0.965926; a calculator gives the cosine of 15 degrees as 0.965926. The graph confirms the size: 15 degrees is small, so its cosine must sit just under one.
\[ \frac{\sqrt6+\sqrt2}{4} = 0.965926 = \cos(15^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 772 — Example 10.4.1 no. 1
This is the payoff for the whole section, and it is the first time you can write an exact value at an angle the unit circle does not list. The only creative step is the first one: deciding which two special angles to combine.
Forty-five minus thirty is the natural choice because both are on the list and you know all four of the values they need. Sixty minus forty-five would also give fifteen and would also work; the answer looks different at first and then simplifies to the same thing, which is a good exercise to try later.
Take the check seriously. Root six plus root two over four comes out as 0.965926, and the graph shows why that is the only plausible answer: fifteen degrees sits between zero and thirty, so its cosine has to sit between one and 0.866. A wrong sign or a swapped term almost always produces a number outside that range, so the check catches it.
Picture it
Figure (svg): A unit circle with rays at 30 and 45 degrees drawn faintly and a ray at 15 degrees drawn boldly, with a dashed line dropping from the 15-degree point to the horizontal axis and a mark on the axis at 0.966.
In degrees, look for a pair from thirty, forty-five, sixty, ninety and their multiples that adds or subtracts to your angle. In radians, read the denominator: a twelve needs a three and a four, because three and four are the denominators whose least common multiple is twelve.
This is the decision that makes these problems easy or hard, and there is a reliable rule for each of the two notations.
In degrees, run through the multiples of thirty and forty-five and look for a pair that adds or subtracts to your angle. Fifteen is forty-five minus thirty; seventy-five is forty-five plus thirty; one hundred and five is sixty plus forty-five. Almost every exam angle is one of these.
In radians, read the denominator instead of the angle. A denominator of twelve is the least common multiple of three and four, so you need one angle with denominator three and one with denominator four. A denominator of twelve almost always means combining thirds and quarters, and once you know that, finding the actual pair is arithmetic with twelfths.
Worked example
The same two special angles, now added rather than subtracted. Watch what happens to the sign.
\[ \text{Find the exact value of } \cos(75^\circ). \]
Write 75 degrees as a sum of special angles
Why: Forty-five plus thirty.
\[ 75^\circ = 45^\circ + 30^\circ \]
Apply the sum identity
Why: Matched terms again, and now the plus in the bracket becomes a minus.
\[ \cos(45^\circ+30^\circ) = \cos 45^\circ\cos 30^\circ - \sin 45^\circ\sin 30^\circ \]
Substitute the four values
Why: Exactly the same four numbers as the previous example.
\[ = \left(\frac{\sqrt2}{2}\right)\left(\frac{\sqrt3}{2}\right) - \left(\frac{\sqrt2}{2}\right)\left(\frac12\right) \]
Multiply each product out
Why: Nothing new; only the sign between them has changed.
\[ = \frac{\sqrt6}{4} - \frac{\sqrt2}{4} \]
Combine over four
Why: One fraction.
\[ = \frac{\sqrt6 - \sqrt2}{4} \]
Check the decimal and the pattern
Why: Root six minus root two over four is 0.258819, and a calculator gives the cosine of 75 degrees as 0.258819. The two answers differ in exactly one sign, because 15 and 75 are the difference and the sum of the same pair of angles.
\[ \cos(75^\circ) = \frac{\sqrt6-\sqrt2}{4} = 0.258819 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 782 — Exercise 7
Do this one yourself before reading the steps. It uses exactly the same two special angles and exactly the same four values as the previous example, so the only thing that can go wrong is the sign, which makes it a clean test of whether the reversal has stuck.
Notice what changed. The bracket now holds a plus, so the answer holds a minus, and the final fraction is root six minus root two over four instead of root six plus root two over four. Nothing else in the calculation moved at all.
The two answers, 0.9659 and 0.2588, are strikingly different even though the working was nearly identical. That is a good argument for slowing down on the sign: it is not a cosmetic detail, it is most of the answer.
Error analysis
A worked solution offered for the cosine of 75 degrees. Two of its three lines are wrong. Find them before revealing.
Annotate
On: \( \cos(75^\circ) = \cos(45^\circ + 30^\circ) = \cos 45^\circ + \cos 30^\circ = \frac{\sqrt2 + \sqrt3}{2} \)
Read the line and try to name both problems before revealing the annotations. One is an error of method and the other is an error you could have caught without knowing any trigonometry at all.
The method error is the distribution the warm-up already ruled out. The other one is size: root two plus root three over two is about 1.573, and no cosine is ever bigger than one. That single observation kills the answer, and it costs about a second.
Make the size check automatic. Every sine and cosine in this course lives between minus one and one, every secant and cosecant is at least one in size, and a tangent can be anything. Whenever you produce a value, glance at it against those bounds before you write it down as final.
Trap
Expanding a cosine of a sum, keeping the sign that was in the bracket:
\[ \cos(\alpha+\beta) \]
\[ = \cos\alpha\cos\beta \]
\[ \quad + \sin\alpha\sin\beta \]
Wrong. The terms are matched correctly, but that is the difference formula.
For cosine the sign always reverses. A plus inside gives a minus outside, and a minus inside gives a plus outside. Sine and tangent keep the bracket's sign; cosine is the one exception.
\[ \cos(\alpha+\beta) \]
\[ = \cos\alpha\cos\beta \]
\[ \quad - \sin\alpha\sin\beta \]
Test any remembered version at alpha and beta both a quarter turn: the cosine of a half turn is minus one, and only the reversing version gives that.
This is the most expensive error in the section, and it is expensive because the rest of the line looks right. The terms are matched correctly, so the working looks like competent work; only one symbol is wrong, and it changes the answer completely.
The defence is a test rather than a rule. Put both angles at a quarter turn, so the sum is a half turn and the cosine of it is minus one. The correct formula gives zero times zero minus one times one, which is minus one. The incorrect one gives plus one. Any remembered version can be checked this way in about five seconds.
It also helps to say out loud what makes cosine different: cosine is the only one of the three whose sign reverses. Sine and tangent both keep the bracket's sign. If you file cosine as the exception rather than as one of three cases, you have one thing to remember instead of three.
Worked example
A verification, not an evaluation. Start on one side and reach the other; never assume the identity you are proving.
\[ \text{Verify: } \cos(\theta - \pi) = -\cos(\theta) \]
Expand the left side with the difference formula
Why: Matched terms, and the minus becomes a plus.
\[ \cos(\theta-\pi) = \cos\theta\cos\pi + \sin\theta\sin\pi \]
Substitute the two values at a half turn
Why: Its cosine is minus one and its sine is zero.
\[ = \cos\theta\,(-1) + \sin\theta\,(0) \]
Simplify
Why: The second term vanishes.
\[ = -\cos\theta \]
Check at a value
Why: At theta equal to 0.9 radians the left side is the cosine of minus 2.241593, which is minus 0.621610, and minus the cosine of 0.9 is also minus 0.621610.
\[ \cos(0.9 - \pi) = -0.621610 = -\cos(0.9) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 783 — Exercise 26
A verification is a different task from an evaluation, and the discipline matters. You may expand, simplify and substitute known values on one side, and you must arrive at the other side. What you may not do is work on both sides at once, or multiply both sides by something, because that assumes the thing you are trying to prove.
Here the left side is the complicated one, so start there. The difference formula turns it into two terms, and both values at a half turn are ones you know: the cosine is minus one and the sine is zero. The second term vanishes and you are done in three lines.
The check at a value is fast and worth the habit. Any old number will do, and if the two sides disagree you have learned that something is wrong before you hand the work in. Agreement at one randomly chosen value is not a proof, but it is strong evidence that your algebra is clean.
Concept
Every identity here is used in both directions, and the right-to-left direction is the one that turns up most in real calculations.
\[ \cos\alpha\cos\beta - \sin\alpha\sin\beta = \cos(\alpha + \beta) \]
\[ \cos\alpha\cos\beta + \sin\alpha\sin\beta = \cos(\alpha - \beta) \]
Left to right you expand one value into four, which is what exact-value problems need. Right to left you collapse four into one, which is what simplification, verification and almost everything in Chapter 11 need.
The signal to watch for is a pair of products built from the same two angles. Check the pairing to decide which family it belongs to, then check the sign to decide whether it is a sum or a difference, and the whole expression becomes a single value.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 772 — Theorem 10.13, read in the other direction
Up to now the identities have been used in one direction: you meet a cosine of a sum and you expand it. That is the direction exact-value problems need, and it is the direction the theorem is written in. It is not the direction most later problems need.
Going the other way, you meet two products in the middle of a longer expression and recognise them as a single value. The signal is always the same shape: two products built from the same pair of angles, added or subtracted. Once you spot it, two decisions finish the job. The pairing tells you which family it belongs to, matched for cosine and mixed for sine, and the sign tells you whether the bracket holds a sum or a difference.
Practise this recognition deliberately, because it does not come for free from being able to expand. Cover the right-hand sides on this slide and try to produce them from the left; that is the skill the next slide tests and the one Chapter 11 assumes.
Prediction
\[ \cos 14^\circ\cos 76^\circ - \sin 14^\circ\sin 76^\circ \]
Predict first
What does this expression equal? No calculator.
Correct: 0
Why: Matched terms with a minus sign between them is the cosine of a SUM, read right to left. The sum is 14 plus 76, which is 90 degrees, and the cosine of a quarter turn is zero. Reading the formulas backwards like this is how they are used in the rest of the chapter.
\[ \cos 14^\circ\cos 76^\circ - \sin 14^\circ\sin 76^\circ = \cos(14^\circ + 76^\circ) = \cos 90^\circ = 0 \]
Commit to an option before revealing. The expression looks like something a calculator is needed for, and the point of the slide is that it is not.
Read the formulas backwards. Two matched products with a minus sign between them is the right-hand side of the cosine sum formula, so the whole expression is the cosine of the sum of the two angles. Fourteen plus seventy-six is ninety, and the cosine of a quarter turn is zero.
This backwards reading is how the identities are mostly used from here on. In Chapter 11 you will meet a pair of products in the middle of a longer calculation and the useful move will be to collapse them into a single cosine or sine. Practise recognising the pattern, not just producing it.
Section
Part 3
Worked example
One line of the difference formula produces the identity that names a whole family, and explains a prefix you have been reading since Section 10.2.
\[ \text{Verify: } \cos\left(\frac{\pi}{2} - \theta\right) = \sin(\theta) \]
Apply the difference formula for cosine
Why: The first angle is a quarter turn; the second is theta.
\[ \cos\left(\frac{\pi}{2}-\theta\right) = \cos\frac{\pi}{2}\cos\theta + \sin\frac{\pi}{2}\sin\theta \]
Substitute the two values at a quarter turn
Why: Its cosine is zero and its sine is one.
\[ = (0)\cos\theta + (1)\sin\theta \]
Simplify
Why: The first term vanishes.
\[ = \sin\theta \]
Check at a value
Why: At theta a sixth of a turn the left side is the cosine of a third of a turn, which is one half, and the sine of a sixth of a turn is also one half.
\[ \cos\left(\frac{\pi}{2}-\frac{\pi}{6}\right) = \cos\frac{\pi}{3} = \frac12 = \sin\frac{\pi}{6} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 773 — Example 10.4.1 no. 2
This is a small calculation with a large consequence. One application of the difference formula, with a quarter turn as the first angle, and two of the four values are zero and one, so almost everything disappears.
What is left is an identity connecting cosine and sine at different angles: the cosine of a quarter turn minus an angle is the sine of that angle. That is the seed of the entire cofunction family, and it also finally explains a piece of vocabulary you have been using since Section 10.2.
Notice how little work it took. The chord derivation was the expensive part; from here the lesson is mostly one-line consequences. Whenever you meet a new formula in this section, it is worth asking which earlier one it came from, because the answer is almost always the cosine difference formula.
Concept
The identity just verified holds for every angle, so it holds for the complement of an angle as well. Feed the complement in and watch the statement fold back on itself.
\[ \sin\left(\frac{\pi}{2} - \theta\right) = \cos\left(\frac{\pi}{2} - \left[\frac{\pi}{2} - \theta\right]\right) = \cos(\theta) \]
So the cosine of an angle is the sine of its complement, and the sine of an angle is the cosine of its complement. Taking reciprocals and quotients of those two carries the pattern to the other four functions.
| this | equals |
|---|---|
| cosine of the complement | sine |
| sine of the complement | cosine |
| secant of the complement | cosecant |
| cosecant of the complement | secant |
| tangent of the complement | cotangent |
| cotangent of the complement | tangent |
Theorem 10.14 — The cofunction identities: for all applicable angles, each of the six functions evaluated at a quarter turn minus the angle equals its cofunction evaluated at the angle.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 773 — Theorem 10.14
The move on this slide is a small trick worth recognising, because it turns up often. You have an identity that holds for every angle, so you are free to feed it any angle you like, including the complement of the one you started with.
Do that and the two quarter turns cancel, leaving the sine of the complement equal to the cosine. So the relationship is symmetric: cosine and sine trade places under complementation. Taking reciprocals of those two statements gives the secant and cosecant versions, and dividing them gives the tangent and cotangent versions, which is why the book leaves the remaining four as exercises.
The table is worth reading as one sentence rather than six rows. Every function of an angle equals its cofunction of the complement. If you know which function is whose cofunction, the table is not something you memorise.
Concept
Cosine, cotangent and cosecant are not arbitrary names. The prefix is short for complement, and it is a claim about angles that this theorem finally makes precise.
Figure (svg): The sine and cosine curves over zero to two pi, with a dashed horizontal line at one half touching the sine curve at pi over six and the cosine curve at pi over three.
\[ \cos\theta = \sin\left(\frac{\pi}{2} - \theta\right) \quad\text{the co-sine is the sine of the co-mplement} \]
It also explains why the cosine graph looks like the sine graph slid along. Replacing the input by a quarter turn minus the input is a reflection followed by a shift, and on a wave that is indistinguishable from a shift alone.
The names were not invented arbitrarily. The prefix is short for complement, and the theorem on the previous slide is the precise statement it was recording. The co-sine of an angle is the sine of its co-mplement, and the same for cotangent and cosecant.
The graph shows the statement as a pair of dots. A sixth of a turn and a third of a turn add to a quarter turn, so they are complements, and the sine at the first equals the cosine at the second. Both are one half, which is why the dashed line touches both curves.
It also explains something about the shapes. Replacing an input by a quarter turn minus the input is a reflection followed by a shift, and on a wave that is indistinguishable from a shift on its own. That is the real reason the cosine graph looks like the sine graph moved along, and you will use it constantly in Section 10.5.
Picture it
Figure (svg): A right triangle with the right angle at the lower right, the angle theta at the lower left and the angle ninety minus theta at the top; the vertical leg is labelled a, the horizontal leg b and the hypotenuse c.
For an acute angle the identity needs no algebra. A right triangle's two acute angles add to a quarter turn, so they are complements, and the leg that is opposite one of them is adjacent to the other. The two corners read the same two ratios with the names swapped.
For an acute angle the cofunction identity needs no algebra at all, and this picture is the reason. The three angles of a triangle add to a half turn, and one of them is a right angle, so the other two add to a quarter turn. The two acute angles of any right triangle are complements.
Now look at what happens to the side labels. Side a is opposite the angle at the bottom left and adjacent to the angle at the top. Side b is the other way round. So the ratio that one corner calls opposite over hypotenuse, the other corner calls adjacent over hypotenuse, and those are the sine and the cosine.
This is worth carrying because it makes the identity feel inevitable rather than remembered. The general theorem covers angles of any size, where there is no triangle to draw, but the triangle is where the idea comes from and where it is easiest to reconstruct under pressure.
Matching
Match the pairs
Why: Each pair of angles adds to a quarter turn, and each pair of names is a function with its cofunction. In degrees look for a total of ninety: 63 with 27, 18 with 72, 25 with 65. In radians a fifth of a turn of half the circle is 36 degrees, whose complement is 54 degrees, which is three tenths of pi.
Make all four pairings before checking. Each one needs two decisions: which angle is the partner, and which function is the partner.
For the angles, look for a total of a quarter turn. In degrees that means adding to ninety, so sixty-three pairs with twenty-seven, eighteen with seventy-two, and twenty-five with sixty-five. In radians it means adding to pi over two, and a fifth of pi is thirty-six degrees, whose complement is fifty-four degrees, which is three tenths of pi.
For the functions, remember that the pairing is a function with its cofunction: sine with cosine, tangent with cotangent, secant with cosecant. The commonest error is getting the angle right and the function wrong, which is why it is worth checking both parts of each answer separately.
Comparison
Comparison matrix
| expression | rewrites as |
|---|---|
| cos(90° − θ) | sin θ |
| sin(90° − θ) | cos θ |
| tan(90° − θ) | cot θ |
| csc(90° − θ) | sec θ |
| cot(90° − θ) | tan θ |
Fill every blank from memory, then check. If you have to look one of them up, you have found the one to practise.
There is only one thing being tested here, and it is whether you know which function is whose cofunction. The angle is the same in every row, so nothing else varies. Sine goes with cosine, tangent with cotangent, secant with cosecant, and each pairing runs in both directions.
Rows three and four are the ones that catch people. The cofunction of the tangent is the cotangent, not the cosine, and the cofunction of the cosecant is the secant, not the sine. The prefix co on the outside of the name is not what decides it.
Trap
A cofunction step applied to the wrong partner:
\[ \sin 70^\circ = \cos 110^\circ \]
Wrong. Seventy and one hundred ten add to a half turn, not a quarter turn.
The cofunction identities pair an angle with its complement, the angle that completes a quarter turn. Seventy pairs with twenty, so the sine of seventy degrees is the cosine of twenty degrees. The supplement is a different relationship entirely, and it is the one behind the sine of a half turn minus theta.
\[ \sin 70^\circ = \cos 20^\circ \]
\[ \sin 70^\circ = \sin 110^\circ \]
Two different relationships are easy to confuse. Complementary angles add to a quarter turn; supplementary angles add to a half turn. The cofunction identities are about complements, and nothing else.
Seventy and one hundred and ten add to one hundred and eighty, so they are supplements, and the cofunction identity has nothing to say about them. The complement of seventy is twenty, so the sine of seventy degrees is the cosine of twenty degrees.
There is a true statement about supplements, and you will prove it later in this lesson as Exercise 27: the sine of a half turn minus an angle equals the sine of that angle. Note that it relates sine to sine, with no change of function at all, which is exactly what makes it a different rule.
Estimation
\[ \frac{\cos 40^\circ}{\sin 50^\circ} \]
Predict first
Estimate this quotient without a calculator.
Correct: exactly 1
Why: Forty and fifty add to ninety, so they are complements, and the cosine of forty degrees is the sine of its complement, which is the sine of fifty degrees. The quotient is a number divided by itself. Seeing this in one glance is what the cofunction identities are for, and it is much faster than computing either value.
\[ \cos 40^\circ = \sin(90^\circ - 40^\circ) = \sin 50^\circ \;\Longrightarrow\; \frac{\cos 40^\circ}{\sin 50^\circ} = 1 \]
Choose an option before revealing, and try to do it without computing either value.
Forty and fifty add to ninety, so they are complements, and the cosine of the first is the sine of the second. The quotient is therefore a number divided by itself, which is one exactly, not approximately. No calculator is involved and no value is ever needed.
Watching for complementary pairs is one of the highest-value habits from this lesson. An expression that looks like it needs two lookups often collapses to a single number, and in longer problems the collapse is what makes the rest of the work possible.
Section
Part 4
Worked example
No new picture. The sine formulas come out of the cosine formula with a cofunction identity at each end.
Turn the sine into a cosine
Why: Read the cofunction identity right to left, with the whole sum as the angle.
\[ \sin(\alpha+\beta) = \cos\left(\frac{\pi}{2} - (\alpha+\beta)\right) \]
Regroup the argument as a difference
Why: Keep the quarter turn with alpha and push beta outside the bracket.
\[ = \cos\left(\left[\frac{\pi}{2}-\alpha\right] - \beta\right) \]
Apply the difference formula for cosine
Why: The first angle is the bracket; the second is beta.
\[ = \cos\left(\frac{\pi}{2}-\alpha\right)\cos\beta + \sin\left(\frac{\pi}{2}-\alpha\right)\sin\beta \]
Apply both cofunction identities
Why: The cosine of the complement is the sine; the sine of the complement is the cosine.
\[ = \sin\alpha\cos\beta + \cos\alpha\sin\beta \]
Get the difference formula by negating beta
Why: Then clear the two negated angles with the even-odd identities.
\[ \sin(\alpha-\beta) = \sin\alpha\cos\beta - \cos\alpha\sin\beta \]
Check at a pair of known angles
Why: Take alpha a third of a turn and beta a sixth, so the sum is a quarter turn, whose sine is one. The right side is three quarters plus one quarter.
\[ \frac{\sqrt3}{2}\cdot\frac{\sqrt3}{2} + \frac12\cdot\frac12 = \frac34 + \frac14 = 1 = \sin\frac{\pi}{2} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 773 — the derivation before Theorem 10.15
This derivation is worth following closely, because it is the pattern the rest of the lesson uses: convert to something you have a formula for, apply the formula, and convert back.
The conversion in both directions is the cofunction identity. Going in, it turns the sine of the sum into a cosine of a quarter turn minus the sum. Coming out, it turns the two complements back into a sine and a cosine of alpha. In between there is a single application of the cosine difference formula.
The regrouping step is where care is needed. The quarter turn has to stay with alpha so that beta ends up outside as a subtraction, which is what makes the cosine difference formula applicable. If you group it the other way you get a true but useless line. Once the sum formula is proved, negating beta gives the difference formula, and there is nothing left to do.
Concept
Theorem 10.15 — For all angles: the sine of a sum is the sine of the first times the cosine of the second plus the cosine of the first times the sine of the second, and the difference version carries a minus in the same place.
\[ \sin(\alpha \pm \beta) = \sin\alpha\cos\beta \pm \cos\alpha\sin\beta \]
Two differences from the cosine formula, and they are the only two things to remember. The terms are mixed rather than matched, and the sign in the answer agrees with the sign in the bracket instead of reversing.
| cosine | sine | |
|---|---|---|
| how the terms pair | matched: cosine with cosine, sine with sine | mixed: sine with cosine, cosine with sine |
| sign in the answer | the opposite of the bracket's | the same as the bracket's |
| symmetric in the two angles? | yes, swapping them changes nothing | no, for the difference the order matters |
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 773 — Theorem 10.15
Two sentences separate the sine formula from the cosine formula, and the table on this slide is those two sentences. The terms are mixed rather than matched, and the sign in the answer agrees with the bracket rather than reversing.
It is worth seeing where each difference comes from. The mixing is the cofunction identities at work: they swapped a sine for a cosine at each end of the derivation, and the swap is what leaves one of each in every term. The sign agreement is simply that no odd function got negated at the wrong moment.
The third row of the table is a detail with consequences. The cosine formula is symmetric in the two angles, so swapping them changes nothing, but the sine difference formula is not. Writing the angles in the wrong order there does not just move a sign, it gives you the negative of the right answer.
Notation
Annotate
On: \( \sin(\alpha + \beta) = \sin\alpha\cos\beta + \cos\alpha\sin\beta \)
Work through the annotations with the cosine formula beside you for comparison, because almost everything here is defined by contrast with it.
The second annotation is the one people skip and then regret. In the sum version the order of the two angles does not matter, so it is easy to stop paying attention to it. In the difference version it matters completely: the first angle keeps the sine in the first term, and swapping them negates the answer.
The last annotation points forward. In Chapter 11, and in every calculus course that touches trigonometry, these formulas are used right to left far more often than left to right. You will be staring at two products and the useful move will be to see them as one sine of a sum. That recognition is a separate skill from expanding, and it is worth practising deliberately.
Worked example
The denominator twelve is the least common multiple of three and four, and that is what tells you which two angles to combine.
\[ \text{Find the exact value of } \sin\left(\frac{19\pi}{12}\right). \]
Write the angle as a sum of special angles
Why: Sixteen twelfths plus three twelfths is nineteen twelfths.
\[ \frac{19\pi}{12} = \frac{4\pi}{3} + \frac{\pi}{4} \]
Apply the sum formula for sine
Why: Mixed terms, and the plus in the bracket stays a plus.
\[ = \sin\frac{4\pi}{3}\cos\frac{\pi}{4} + \cos\frac{4\pi}{3}\sin\frac{\pi}{4} \]
Substitute the four values
Why: Four thirds of pi lands in quadrant three, where both coordinates are negative.
\[ = \left(-\frac{\sqrt3}{2}\right)\left(\frac{\sqrt2}{2}\right) + \left(-\frac12\right)\left(\frac{\sqrt2}{2}\right) \]
Multiply each product out
Why: Two negative terms.
\[ = -\frac{\sqrt6}{4} - \frac{\sqrt2}{4} \]
Combine over four
Why: One fraction.
\[ = \frac{-\sqrt6 - \sqrt2}{4} \]
Figure (svg): A unit circle with faint rays at 240 and 45 degrees and a bold ray at 285 degrees reaching a point low in the fourth quadrant, with a dashed segment from that point across to the vertical axis marked at minus 0.966.
Check the sign and the decimal
Why: Nineteen pi over twelve is 285 degrees, which lies in quadrant four, where the sine is negative and close to minus one. The value is minus 0.965926, and a calculator agrees.
\[ \frac{-\sqrt6-\sqrt2}{4} = -0.965926 = \sin(285^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 774 — Example 10.4.2 no. 1
This one tests the decomposition rule in radians. Read the denominator, not the angle: a twelve means you need a three and a four, so you are combining an angle with denominator three and an angle with denominator four.
Converting to twelfths makes the search easy. Sixteen twelfths is four thirds of pi and three twelfths is a quarter of pi, and they add to nineteen twelfths. Four thirds of pi is in the third quadrant, where both coordinates are negative, which is why both terms come out negative.
The check is doing real work here, not decoration. Nineteen twelfths of pi is two hundred and eighty-five degrees, which the picture shows landing deep in the fourth quadrant, just short of the bottom of the circle. So the sine must be negative and close to minus one, and minus 0.9659 is exactly that. If your answer had come out positive, the picture would have caught it.
Step zero
\[ \sin\alpha = \frac{5}{13} \;(\text{quadrant II}), \qquad \tan\beta = 2 \;(\text{quadrant III}) \]
Discussion prompt
The next example asks for the sine of alpha minus beta from exactly this information. Before touching the formula, list every quantity the formula will demand and mark which ones you have not been given.
Write your list before revealing. Thirty seconds of planning here saves the usual disaster, which is getting three quarters of the way through and realising a sign was never decided.
The formula needs four quantities and you have been handed one of them. The cosine of alpha comes from the Pythagorean identity, which gives it up to sign, and the quadrant decides the sign. The two quantities for beta both have to be built from its tangent, and again the quadrant decides the signs.
Notice that the quadrant information is not decoration in the problem statement; it is load-bearing. Without it, each of the three missing quantities would have two possible values, and the final answer would have eight. Every time a problem tells you a quadrant, it is telling you which of two square roots to take.
Worked example
Neither angle is named. Both terminal points must be rebuilt before the formula can be used.
\[ \sin\alpha = \tfrac{5}{13} \text{ in QII}, \quad \tan\beta = 2 \text{ in QIII}; \quad \text{find } \sin(\alpha - \beta). \]
Find the cosine of alpha from the Pythagorean identity
Why: Twelve over thirteen, up to sign.
\[ \cos^2\alpha = 1 - \left(\frac{5}{13}\right)^2 = \frac{144}{169} \;\Longrightarrow\; \cos\alpha = \pm\frac{12}{13} \]
Choose the sign from the quadrant
Why: In quadrant two the first coordinate is negative.
\[ \cos\alpha = -\frac{12}{13} \]
Get the secant of beta from the tangent Pythagorean identity
Why: One plus the tangent squared is the secant squared.
\[ \sec^2\beta = 1 + 2^2 = 5 \;\Longrightarrow\; \sec\beta = \pm\sqrt5 \]
Choose the sign and invert to reach the cosine
Why: Quadrant three has a negative first coordinate, so the secant is negative.
\[ \sec\beta = -\sqrt5 \;\Longrightarrow\; \cos\beta = -\frac{1}{\sqrt5} = -\frac{\sqrt5}{5} \]
Get the sine of beta from the quotient identity
Why: The tangent is the sine over the cosine, so the sine is the tangent times the cosine.
\[ \sin\beta = \tan\beta\cos\beta = 2\left(-\frac{\sqrt5}{5}\right) = -\frac{2\sqrt5}{5} \]
Figure (svg): A unit circle with a ray into the second quadrant reaching the point minus twelve thirteenths, five thirteenths, and a ray into the third quadrant reaching the point minus root five over five, minus two root five over five.
Substitute into the difference formula for sine
Why: Mixed terms, and the minus in the bracket stays a minus.
\[ \sin(\alpha-\beta) = \left(\frac{5}{13}\right)\left(-\frac{\sqrt5}{5}\right) - \left(-\frac{12}{13}\right)\left(-\frac{2\sqrt5}{5}\right) \]
Multiply both products out
Why: The second product is positive, and the minus in front of it survives.
\[ = -\frac{\sqrt5}{13} - \frac{24\sqrt5}{65} \]
Combine over sixty-five
Why: Root five over thirteen is five root five over sixty-five.
\[ = -\frac{5\sqrt5}{65} - \frac{24\sqrt5}{65} = -\frac{29\sqrt5}{65} \]
Check the size and the sign
Why: The value is minus 0.997630, and a sine can never exceed one in size, so it is at least possible. Computing the angles numerically gives alpha 2.746802 and beta 4.248741 radians, and the sine of their difference is minus 0.997630.
\[ -\frac{29\sqrt5}{65} = -0.997630 = \sin(2.746802 - 4.248741) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 774 — Example 10.4.2 no. 2
This is the hardest kind of problem in the lesson, and it is hard for bookkeeping reasons rather than conceptual ones. The identity itself is one substitution; everything before it is reconstruction.
Work in the order the steps use. For alpha, the Pythagorean identity gives the cosine up to sign and the quadrant settles it. For beta, the tangent version of the Pythagorean identity is the quick route to the secant, hence to the cosine, and then the quotient identity gives the sine as the tangent times the cosine. The book notes that you could use the ordinary Pythagorean identity instead; it chooses the quotient identity because it is shorter.
The picture is there so that the signs stop being arithmetic and become geometry. Alpha's point is up and to the left, beta's is down and to the left, and every sign in the final substitution can be read straight off the two points. Then check the size: a sine of minus 0.9976 is only just possible, and the numerical evaluation of the two angles confirms it.
Error analysis
An expansion of the sine of a difference, offered on a test. Two things are wrong with it.
Annotate
On: \( \sin(\alpha - \beta) = \sin\alpha\sin\beta - \cos\alpha\cos\beta \)
Try to name the problem before revealing. There are two things to notice, and the second one is a technique you should adopt permanently.
The mistake is that the terms have been matched, sine with sine and cosine with cosine. That is the cosine formula's pattern. A sine formula always mixes: each of its two terms has exactly one sine and one cosine. If you ever write a product of two sines inside a sine expansion, something has gone wrong.
The technique is the test in the second annotation. Set beta to zero. The left side collapses to the sine of alpha, and any correct formula must reproduce that. The offered line gives minus the cosine of alpha instead, which is wrong for every value, so it is not an identity. One substitution, five seconds, and it exposes almost every mis-remembered version of these formulas.
Worked example
A verification whose consequence you will meet again in Section 11.2, when a given sine leaves two possible triangles.
\[ \text{Verify: } \sin(\pi - \theta) = \sin(\theta) \]
Expand with the difference formula for sine
Why: Mixed terms, and the minus in the bracket stays a minus.
\[ \sin(\pi - \theta) = \sin\pi\cos\theta - \cos\pi\sin\theta \]
Substitute the two values at a half turn
Why: Its sine is zero and its cosine is minus one.
\[ = (0)\cos\theta - (-1)\sin\theta \]
Simplify
Why: Two minus signs make a plus.
\[ = \sin\theta \]
Figure (svg): A unit circle with rays at forty and one hundred forty degrees reaching two points at the same height, joined by a dashed horizontal line, with the shared height marked on the vertical axis.
Check at a value and read the consequence
Why: At theta equal to 0.9 radians both sides are 0.783327. In the picture the two terminal points sit at the same height, which is why a sine value on its own can never name its angle.
\[ \sin(\pi - 0.9) = 0.783327 = \sin(0.9) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 783 — Exercise 27
The verification itself is three lines and uses the same discipline as the earlier one: expand the complicated side and drive it to the simple one. The values at a half turn are a sine of zero and a cosine of minus one, and the two minus signs in the second term cancel.
The picture is the part to remember. Two different angles, forty degrees and one hundred and forty degrees, have terminal points at exactly the same height, so they have exactly the same sine. Their cosines are opposite, but their sines are identical.
That is the whole reason the law of sines has an ambiguous case, which you will meet in Section 11.2. If a triangle problem tells you a sine and asks for the angle, there are two candidate angles, and only the rest of the problem can decide between them. Knowing the identity now means the ambiguity will look inevitable rather than surprising.
Fill the middle
Add the sum and difference formulas for sine and watch half the terms disappear. This is the first of the product-to-sum formulas, which Lesson 10.4c builds on.
\[ \sin(\alpha+\beta) = \sin\alpha\cos\beta + \cos\alpha\sin\beta \]
\[ \sin(\alpha-\beta) = \sin\alpha\cos\beta - \cos\alpha\sin\beta \]
Fill in the blanks
\sin(\alpha+\beta) + \sin(\alpha-\beta) = 2\,\sin(\alpha)\,\cos(\beta)
Why: Adding the two lines makes the cosine-alpha sine-beta terms cancel, because one carries a plus and the other a minus. What survives is the sine-alpha cosine-beta term twice. At alpha 0.7 and beta 0.3 both sides equal 1.230889, so the identity checks numerically as well.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 783 — Exercise 29
Fill both blanks before checking. The two formulas you need are printed above, so this is a question about what cancels rather than about what the formulas are.
Line them up and add. The first term is the same in both, so it appears twice. The second term appears once with a plus and once with a minus, so it disappears entirely. What survives is twice a product of one sine and one cosine, at different angles.
This identity is more important than it looks. It converts a sum of two sines into a product, and Lesson 10.4c is built on exactly this kind of conversion. It is also the reason the phase-shift slider two slides from now behaves the way it does, so it is worth having in hand.
Tweak it
Parameter explorer
The curve is the sine wave plus a copy of itself delayed by p. Slide p from zero upward. The wavelength never changes; only the height does. At what value of p does the sum vanish completely?
\[ \sin x + \sin(x + {p}) \]
\[ \sin x + \sin(x+p) = 2\cos\left(\frac{p}{2}\right)\sin\left(x + \frac{p}{2}\right) \]
Start with the slider near zero and move it up slowly. Watch two things separately: the wavelength, which never changes at all, and the height, which shrinks steadily.
The wavelength staying fixed is the real content. Adding two waves of the same frequency can only ever give you a wave of that same frequency, no matter what the phase gap is. That is not obvious from the picture and it is not obvious from physical intuition, but the sum identity forces it.
The height is governed by the factor on the slide: twice the cosine of half the phase gap. At a gap of zero that is two, which is full reinforcement. Keep pushing and it falls to zero when the gap reaches half a turn, at which point the two waves cancel completely and the curve flattens onto the axis. That is exactly the mechanism a noise-cancelling headphone uses.
Section
Part 5
Worked example
Nothing new is proved here. The tangent formula is the quotient of the two formulas already in hand, tidied so that only tangents appear.
Write the tangent as a quotient
Why: The quotient identity, at the angle alpha plus beta.
\[ \tan(\alpha+\beta) = \frac{\sin(\alpha+\beta)}{\cos(\alpha+\beta)} \]
Expand the top and the bottom
Why: Sine of a sum above, cosine of a sum below.
\[ = \frac{\sin\alpha\cos\beta + \cos\alpha\sin\beta}{\cos\alpha\cos\beta - \sin\alpha\sin\beta} \]
Divide every one of the four terms by the product of the two cosines
Why: Dividing the top and the bottom by the same nonzero quantity changes nothing.
\[ = \frac{\dfrac{\sin\alpha\cos\beta}{\cos\alpha\cos\beta} + \dfrac{\cos\alpha\sin\beta}{\cos\alpha\cos\beta}}{\dfrac{\cos\alpha\cos\beta}{\cos\alpha\cos\beta} - \dfrac{\sin\alpha\sin\beta}{\cos\alpha\cos\beta}} \]
Cancel inside each of the four fractions
Why: Each one leaves a single tangent, or a one.
\[ = \frac{\tan\alpha + \tan\beta}{1 - \tan\alpha\tan\beta} \]
Get the difference formula by negating beta
Why: The tangent is odd, so both occurrences of the tangent of beta change sign.
\[ \tan(\alpha-\beta) = \frac{\tan\alpha - \tan\beta}{1 + \tan\alpha\tan\beta} \]
Check at a pair of known angles
Why: Take both angles a sixth of a turn, where each tangent is root three over three. The formula gives two root three over three divided by two thirds, which is root three, and the tangent of a third of a turn is root three.
\[ \frac{\frac{\sqrt3}{3}+\frac{\sqrt3}{3}}{1 - \frac13} = \frac{\frac{2\sqrt3}{3}}{\frac23} = \sqrt3 = \tan\frac{\pi}{3} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 775-776 — Example 10.4.2 no. 3 and the remark after it
Nothing new is proved here, and it is worth being clear about that before you start. The tangent formula is an algebraic tidy-up of two formulas you already have.
The one clever step is dividing every term, top and bottom, by the product of the two cosines. Why that and not something else? Because each of the four terms contains at most one sine and at most one cosine from each angle, so dividing by the product of the cosines turns every term into either a tangent or a one. Nothing else has that effect.
Two cautions come out of the derivation. The division assumed both cosines were nonzero, which is why the formula is silent when either angle is a quarter turn. And the difference version comes from negating beta, which flips the tangent of beta in two places at once, so the numerator's sign changes and so does the denominator's.
Concept
The book compresses the six formulas into three, using a pair of stacked signs. The top sign of every symbol belongs to the sum; the bottom sign of every symbol belongs to the difference.
\[ \cos(\alpha \pm \beta) = \cos\alpha\cos\beta \mp \sin\alpha\sin\beta \]
\[ \sin(\alpha \pm \beta) = \sin\alpha\cos\beta \pm \cos\alpha\sin\beta \]
\[ \tan(\alpha \pm \beta) = \frac{\tan\alpha \pm \tan\beta}{1 \mp \tan\alpha\tan\beta} \]
Read the cosine line downwards and you get the two cosine formulas. The upside-down sign in the cosine line, and in the tangent's denominator, is exactly the sign reversal that has been the theme of this lesson.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 775-776 — Theorem 10.16
The stacked signs are a compression device, not a new idea. Six formulas have been written as three, and the rule for reading them is simple: choose the sum and take the top of every stacked symbol; choose the difference and take the bottom of every one.
Read the cosine line both ways and check that you recover the two formulas you already know. Then do the same with the tangent line, and notice that the two stacked symbols in it point in opposite directions, so choosing the sum gives a plus above and a minus below.
The upside-down symbols are exactly where the sign reversal lives, and the reversal has been the theme of the whole lesson. It appears in the cosine formula and in the tangent's denominator, and nowhere else. If you can point at those two places, you have the signs of all six formulas.
Notation
Annotate
On: \( \tan(\alpha \pm \beta) = \frac{\tan\alpha \pm \tan\beta}{1 \mp \tan\alpha\tan\beta} \)
Work through the annotations slowly, because notation is the whole difficulty on this slide. Nothing here is new mathematics; it is a question of reading two formulas out of one line without mixing them.
The most practical annotation is the last one. Under exam pressure, do not try to hold both sign choices in your head while also doing the algebra. Write the version you need out in full first, with real plus and minus signs, and then start substituting. The compressed form is for reference, not for working in.
The third annotation is easy to forget once the formula looks clean. Because every sine and cosine was divided away, nothing on the right reminds you that the derivation needed both cosines to be nonzero. The counterexample slide coming up makes that restriction concrete.
Picture it
Figure (svg): Two lines through the origin, one of slope one half and one of slope three, with an arc marking the forty-five degree angle between them and the formula for the tangent of that angle written beside it.
A line through the origin has slope equal to the tangent of the angle it makes with the positive horizontal direction. So the angle between two lines is a difference of two such angles, and the tangent difference formula computes its tangent from the two slopes alone. Here the slopes are one half and three, the formula returns one, and the angle between the lines is therefore forty-five degrees.
This is the tangent difference formula doing geometry, and it is worth seeing because it explains why the formula has the shape it does rather than just asserting it.
Start from the fact that a line through the origin has slope equal to the tangent of the angle it makes with the positive horizontal direction. So two lines correspond to two angles, and the angle between the lines is the difference of those angles. The tangent of that difference is therefore computable from the two slopes alone, with no angles in sight.
The numbers here are chosen to be memorable. The slopes are one half and three; the formula gives two and a half over two and a half, which is exactly one; so the angle between the lines is forty-five degrees on the nose. Try it with a few other pairs of slopes and you have a tool for a whole family of geometry problems.
Worked example
The same denominator rule as before, now feeding the tangent formula, which needs only two values instead of four.
\[ \text{Find the exact value of } \tan\left(\frac{13\pi}{12}\right). \]
Write the angle as a difference of special angles
Why: Sixteen twelfths minus three twelfths is thirteen twelfths.
\[ \frac{13\pi}{12} = \frac{4\pi}{3} - \frac{\pi}{4} \]
Apply the difference formula for tangent
Why: Minus on the top, plus underneath.
\[ = \frac{\tan\frac{4\pi}{3} - \tan\frac{\pi}{4}}{1 + \tan\frac{4\pi}{3}\tan\frac{\pi}{4}} \]
Substitute the two tangents
Why: Four thirds of pi has tangent root three; a quarter of pi has tangent one.
\[ = \frac{\sqrt3 - 1}{1 + \sqrt3} \]
Rationalise the denominator
Why: Multiply the top and the bottom by root three minus one.
\[ = \frac{(\sqrt3-1)^2}{(\sqrt3+1)(\sqrt3-1)} = \frac{4 - 2\sqrt3}{2} \]
Divide both terms by two
Why: Both are even.
\[ = 2 - \sqrt3 \]
Check the decimal
Why: Two minus root three is 0.267949. Thirteen pi over twelve is 195 degrees, which is a half turn plus 15 degrees, so its tangent equals the tangent of 15 degrees, and a calculator gives 0.267949.
\[ 2 - \sqrt3 = 0.267949 = \tan(195^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 782 — Exercise 15
The tangent formula is often the fastest of the three, because it needs two values instead of four. The cost is an extra step at the end, since the answer arrives as a fraction with a radical underneath.
The decomposition rule is unchanged: a denominator of twelve means combining thirds and quarters, and sixteen twelfths minus three twelfths is thirteen twelfths. Then the difference version of the formula has a minus on top and a plus underneath, which is the reversal to watch.
Rationalising is worth doing carefully. Multiplying top and bottom by root three minus one squares the numerator and turns the denominator into a plain two, and the answer simplifies to two minus root three. The check is satisfying: thirteen twelfths of pi is one hundred and ninety-five degrees, which is a half turn past fifteen degrees, and the tangent repeats every half turn, so the answer had to be the tangent of fifteen degrees.
Concept
Several exercises in this section ask for a formula at an angle shifted by a quarter turn or a half turn. They are all one calculation, because one of the two angles is special, so most of the four products vanish.
\[ \cos(\theta - \pi) = -\cos\theta, \qquad \sin(\pi - \theta) = \sin\theta \]
At a quarter turn the cosine is zero and the sine is one; at a half turn the cosine is minus one and the sine is zero. Either way two of the four products disappear and what survives is a single function of theta, sometimes negated.
The tangent case needs one extra thought, because the tangent of a quarter turn does not exist and the tangent sum formula is therefore unavailable. Expand the sine and the cosine separately and divide them instead.
\[ \tan\left(\theta + \frac{\pi}{2}\right) = \frac{\sin\left(\theta+\frac{\pi}{2}\right)}{\cos\left(\theta+\frac{\pi}{2}\right)} = \frac{\cos\theta}{-\sin\theta} = -\cot\theta \]
So there is no separate list of reduction formulas to learn. Expand, substitute the two special values, and the answer is three lines away.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 783 — Exercises 26, 27 and 28
Some textbooks hand you a page of reduction formulas to memorise: what happens to each function when the angle is shifted by a quarter turn, a half turn or three quarters of a turn. You do not need that page, and this slide is why.
Every one of them is a sum or difference formula with a special angle in one slot. At a quarter turn the cosine is zero and the sine is one; at a half turn the cosine is minus one and the sine is zero. Two of the four products are therefore multiplied by zero and disappear, and the survivor is a single function of the remaining angle, sometimes with a minus sign attached.
The tangent case is the one worth remembering separately, and only because of a restriction. The tangent of a quarter turn does not exist, so the tangent sum formula cannot be used at all. Go back to the definition, expand the sine and the cosine separately, and divide. The cosine on top over the negated sine underneath is the negative of the cotangent, and the whole thing takes three lines.
Trap
The tangent of a difference, with the bracket's sign copied everywhere:
\[ \tan(\alpha-\beta) \]
\[ = \frac{\tan\alpha - \tan\beta}{1 - \tan\alpha\tan\beta} \]
Wrong underneath. The numerator is right; the denominator reverses.
In the tangent formula the numerator takes the bracket's sign and the denominator takes the opposite one. Test it the same way you test everything else: put both angles at a sixth of a turn, where the difference is zero and the tangent must be zero.
\[ \tan(\alpha-\beta) = \frac{\tan\alpha - \tan\beta}{1 + \tan\alpha\tan\beta} \]
\[ \frac{0}{1 + \frac13} = 0 = \tan 0 \]
This is the same sign-reversal error as before, relocated. The numerator of the tangent formula takes the bracket's sign, exactly as in the sine formula, so it is natural to assume the denominator does too. It does not.
Fix it from the derivation rather than from memory. The tangent formula is the sine formula over the cosine formula, and the cosine formula is the one whose sign reverses. So the reversal lands in the denominator, because that is where the cosine formula ended up.
The test on the right is the cheapest one available. Put both angles equal, so the difference is zero and the tangent must be zero. The correct formula gives a numerator of zero and a harmless denominator, which is zero. A version with the wrong denominator sign still gives zero here, so use a pair of unequal angles as well: at a third and a sixth of a turn the difference is a sixth of a turn, whose tangent is root three over three.
Counterexample
A claim worth attacking: the tangent sum formula works for any two angles at all.
Discussion prompt
Find two angles for which the right-hand side cannot be evaluated, and say whether the left-hand side is in trouble too. Then find two angles where the right-hand side is fine but the left-hand side is not.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 775 — the closing remark of Example 10.4.2 no. 3
Try both parts before revealing, and notice that they fail in genuinely different ways. This is not a slide about a formula being wrong; it is about the difference between a formula being false and a formula being inapplicable.
In the first case both sides break together. The denominator is zero and the tangent of a quarter turn does not exist, so the formula is describing a failure that is really there. Nothing is contradicted.
The second case is the interesting one. The left side is a perfectly ordinary minus one, while the right side cannot even be written down, because one of the tangents it needs does not exist. So the formula is not equivalent to the statement it came from: it is narrower. The book's closing remark, that the formula is limited to cases where all the tangents are defined, is exactly this observation, and when an angle is a quarter turn you fall back on the sine and cosine formulas, which have no such restriction.
Comparison
Comparison matrix
| formula | how the terms pair | sign in the answer |
|---|---|---|
| cosine of a sum | matched | minus |
| cosine of a difference | matched | plus |
| sine of a sum | mixed | plus |
| sine of a difference | mixed | minus |
| tangent of a sum | tangents only | plus on top, minus below |
| tangent of a difference | tangents only | minus on top, plus below |
Fill every blank from memory before checking. Six formulas is a lot to hold, and this table is arranged so that you do not have to: each row differs from its neighbours in one place only.
Read down the middle column first. Cosine matches its terms, sine mixes them, and tangent has no sines or cosines left at all. That is three facts, and they never vary.
Then read down the right-hand column, which is where all the errors live. Cosine reverses. Sine agrees. Tangent agrees on top and reverses underneath. If you can say those three sentences, every sign in the lesson follows, and the last two rows of the table are just the tangent sentence written out.
Section
Part 6
Intuition
A calculator returns the cosine of 15 degrees as 0.9659258 in one keystroke. So why spend six lines producing root six plus root two, all over four?
Because a decimal is a dead end and an exact value is not. Exact values cancel against each other, combine into further exact values, and let you see that an answer is a particular number rather than merely close to one.
\[ \frac{\sqrt6+\sqrt2}{4}\cdot\frac{\sqrt6-\sqrt2}{4} = \frac{6-2}{16} = \frac14 \]
That product is the cosine of 15 degrees times the cosine of 75 degrees, and it is exactly one quarter. The same computation in decimals gives 0.2499999 and leaves you guessing whether the pattern is real.
It is the same reason calculus, physics and engineering keep answers symbolic for as long as possible. Structure is visible in the symbols and invisible in the digits.
It is a fair question, and it deserves a real answer rather than because the exam says so. A calculator gives the cosine of 15 degrees instantly, so the six lines of algebra look like a ritual.
The answer is that exact values keep working and decimals stop. Look at the product on the slide: the cosine of 15 degrees times the cosine of 75 degrees. In exact form the radicals multiply out as a difference of two squares and the answer is one quarter, on the nose. In decimals you get 0.2499999 and you have no way to tell whether that is one quarter or merely near it.
This matters far beyond this lesson. Almost every simplification in the rest of trigonometry depends on terms cancelling exactly, and a term that has been rounded cannot cancel. The same reasoning is why a physicist carries symbols through a derivation and substitutes numbers only at the very end: the structure of the answer is visible in the symbols and completely hidden in the digits.
Pattern
Figure (svg): A flow diagram: read what the angle is made of, then branch three ways to an exact value by decomposing into special angles, to a verification by expanding one side, or to an evaluation from given values and quadrants.
This is the order to work in, whatever the question looks like, and the first item is the one most often skipped.
Clearing minus signs from inside arguments comes before anything else, because a hidden factor of minus one is the usual reason a problem does not match any formula you know. After that, the question type decides the route: an exact value means splitting the angle, a verification means expanding one side only, and a problem stated with values and quadrants means rebuilding both terminal points before you substitute.
The last item is not optional. Two cheap checks catch most errors: any sine or cosine outside minus one to one is impossible, and setting one angle to zero tests a remembered formula instantly. Both take seconds and both have appeared on this lesson's error-analysis slides for a reason.
Ranking
Put in order
This whole lesson is one theorem and five consequences. Put the derivations in the order the book proves them.
Why: The even-odd identities come first because the chord proof and every later step use them to absorb a negated angle. The chord argument is the only geometry. The sum case for cosine is then a one-line consequence, the cofunction identities are a special case of that, the sine formulas need the cofunction identities, and the tangent formula needs both sine and cosine. Take any link out and what follows it collapses.
Put them in order before checking. The point is not the order for its own sake; it is to see that this lesson contains one theorem and five consequences, which is far less to learn than five theorems.
The even-odd identities have to come first, because the chord proof uses them to handle the case where the second angle is larger, and every later derivation uses them to absorb a negated angle. Then comes the only piece of geometry in the section. Everything after that is algebra on what is already proved.
It is worth tracing the dependencies once. Remove the cofunction identities and you lose the sine formulas, because the sine derivation goes through them. Remove the sine formulas and you lose the tangent formula, because it is their quotient with the cosine ones. That is why the order is not negotiable, and why a forgotten formula can always be rebuilt from the ones before it.
Check
Check your understanding
Which of these is NOT an identity?
Answer: C
Why: Cotangent is cosine over sine. Negating the angle leaves the cosine alone and negates the sine, so the quotient changes sign: the cotangent is odd, and the correct identity carries a minus sign on the right.
Three of these four are genuine identities and one is not, so the question is really whether you can produce each of the four from scratch rather than recall which had a minus sign.
Write each function as cosine and sine and count the sine factors. Secant has none, so it is even. Cosecant and tangent have one each, so both are odd. Cotangent also has one, in its denominator, so it is odd too, and the option that claims otherwise is the false one.
If you were tempted by the cotangent option, the likely cause is the prefix co, which suggests a partnership with the cosine. There is such a partnership, but it is the cofunction relationship, not evenness. Cotangent is odd.
Check
Check your understanding
Which expansion of cos(α + β) is correct?
Answer: B
Why: The cosine formula matches its terms, cosine with cosine and sine with sine, and reverses the sign: a plus inside the bracket gives a minus in the answer. Testing both angles at a quarter turn gives minus one on the left, which only this option reproduces.
All four options are things people write. Two of them are correct formulas for something else, one is distribution, and one is the answer.
Use the test rather than memory: put both angles at a quarter turn. The left side becomes the cosine of a half turn, which is minus one. Only the option with matched terms and a minus sign gives that, because the other matched option gives plus one and the mixed option gives zero.
If you chose the first option, you have the cosine difference formula and the sign reversal is what slipped. That is the single most common error in this section, and the fix is to file cosine as the one formula whose sign flips.
Check
Check your understanding
Which of these equals sin 63°?
Answer: B
Why: The cofunction identity pairs an angle with its complement, and 63 plus 27 is 90. So the sine of 63 degrees is the cosine of 27 degrees; both are 0.891007.
Two decisions are being tested at once: which angle is the partner, and which function is the partner. Getting one right and the other wrong is the usual failure, and three of the four options are built from exactly that.
The partner angle is the complement, so sixty-three pairs with twenty-seven, not with one hundred and seventeen. The partner function is the cofunction, so the sine of an angle equals the cosine of its complement, not the sine of its complement. Only one option gets both right, and both sides come out as 0.891007.
If you picked the supplement option, notice that it is not even the right sign: the cosine of one hundred and seventeen degrees is negative while the sine of sixty-three degrees is positive. A quick quadrant glance would have ruled it out before any identity was needed.
Real world
Figure (svg): Three curves over zero to twelve: two sine waves of amplitude one, one shifted right by about one unit, and their sum, a sine wave of the same wavelength with amplitude about 1.76 marked by a dashed line.
Discussion prompt
Two speakers play the same pure tone, but one signal reaches your ear slightly later, so what you hear is the sine of a quantity plus the sine of that quantity plus a fixed phase gap. Use the sum identity to explain why the result is still a tone of the same pitch, and say what decides whether the two reinforce or cancel.
OpenStax Algebra and Trigonometry 2e, §9.2 Sum and Difference Identities §9.2 — the same identities, with engineering applications
Write your explanation before revealing. The physical setting is familiar, and the algebra on the slide is a single application of the sum identity, so this is mostly a question of connecting the two.
The key claim is that the pitch does not change. Expanding the delayed sine with the sum identity and collecting terms leaves a single sine whose argument runs at the same rate as before. Only two things have changed: the wave has slid along by half the phase gap, and its amplitude has been multiplied by twice the cosine of half the phase gap.
That amplitude factor is the whole story of interference. With no delay it is two, and the tones reinforce. As the delay grows toward half a period the factor falls to zero, and the tones cancel completely. Noise-cancelling headphones engineer that cancellation deliberately; the dead spots in a badly wired room produce it by accident. Either way, the explanation is one trigonometric identity.
Explain it to yourself
Discussion prompt
In four or five sentences, and without looking back, explain how the sine sum formula follows from the cosine difference formula. Name every identity you use along the way.
Write the explanation out before revealing, in sentences rather than symbols. Being able to say the chain out loud is the difference between owning these formulas and renting them.
The route has three stages and each uses a named identity. The cofunction identity converts the sine into a cosine. The cosine difference formula expands it. The cofunction identities convert the two complements back. If you can name all three, you can rebuild the sine formula at any point in the future from the cosine one.
There is a bonus in doing this. The mixing of terms, which is otherwise just a fact to memorise, becomes explicable: the cofunction identities swap a sine for a cosine at each end of the derivation, and that swap is precisely what leaves one of each in every term. A formula you can explain is a formula you will not misremember.
Edge cases
Exercise 36 in the section's exercise set does something these formulas were partly invented for. It is the first step of every derivative you will meet in a trigonometry-heavy calculus course.
\[ \frac{\sin(t+h) - \sin(t)}{h} \]
Discussion prompt
Expand the top with the sum identity and rearrange until the expression is a combination of the sine of t and the cosine of t whose coefficients involve h but not t. Then say what each coefficient has to do as h shrinks, if the derivative of the sine is to be the cosine.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 783 — Exercise 36
Have a go before revealing. The algebra is short, and the payoff is seeing why anyone would care about these formulas beyond exact values at odd angles.
Expand the top with the sum identity and two of the three resulting terms contain the sine of t. Cancel and regroup so that the sine of t and the cosine of t each sit outside a bracket containing only h. The variables have now been separated, which is the point of the whole manoeuvre.
Look at what has to happen next. For the whole expression to approach the cosine of t as h shrinks, the first bracket must approach one and the second must approach zero. Those two limits are exactly what a calculus course proves first about the trigonometric functions, and the sum identity is what reduced an unfamiliar derivative to them. This is the reason the identity is worth having as a formula rather than as a curiosity.
Exit ticket
\[ \text{Find the exact value of } \sin(105^\circ), \text{ then simplify } \frac{\cos(-\theta)}{\sin\left(\frac{\pi}{2}-\theta\right)}. \]
Discussion prompt
Do both, and for each one name the identity you used and the reason you chose that decomposition or that rewrite.
Do both parts on paper before revealing, and for each one write down the identity you used and why you chose it. The naming is half the exercise.
For the first part, one hundred and five degrees splits as sixty plus forty-five, so the sum formula for sine applies with values you know. The answer is root six plus root two over four, which is 0.965926, the same number as the cosine of fifteen degrees. That is not a coincidence: one hundred and five and fifteen are complements, and the cofunction identity says their sine and cosine must agree.
For the second part, the numerator needs the even identity for cosine and the denominator needs the cofunction identity for sine. Both come out as the cosine of theta, so the quotient is one wherever that cosine is nonzero. If you can name which identity did which job, you have the lesson.
Recap
| identity | the statement | where it came from |
|---|---|---|
| even-odd | cosine and secant ignore a negated angle; the other four change sign | reflection across the horizontal axis |
| cosine of a sum | matched terms, and the answer's sign reverses | two equal chords, then the distance formula |
| cofunction | each function of an angle is its cofunction of the complement | the cosine difference at a quarter turn |
| sine of a sum | mixed terms, and the answer's sign agrees | the cosine formula via a cofunction identity |
| tangent of a sum | tangents only, denominator's sign reversed | sine formula divided by cosine formula |
\[ \cos(\alpha\pm\beta) = \cos\alpha\cos\beta \mp \sin\alpha\sin\beta \]
\[ \sin(\alpha\pm\beta) = \sin\alpha\cos\beta \pm \cos\alpha\sin\beta \]
Next, Lesson 10.4b specialises the sum formulas to the case where the two angles are equal, which collapses them into the double angle identities and from there into the power reduction and half angle formulas.
OpenStax Algebra and Trigonometry 2e, §9.1 Verifying Trigonometric Identities and Using Trigonometric Identities to Simplify Trigonometric Expressions §9.1 — verifying identities, with more practice problems
Five identities, and only one of them was proved from a picture. The chord argument gave the cosine of a difference; the sum version came from negating an angle; the cofunction identities came from putting a quarter turn into the difference formula; the sine formulas came from routing through the cofunction identities; and the tangent formula came from dividing. Remember the chain and you never depend on remembering a sign.
If you take away two sentences, take these. The cosine formula matches its terms and reverses its sign. The sine formula mixes its terms and keeps its sign. The tangent formula follows from those two, with the reversal landing in the denominator.
Next comes the special case where the two angles are equal. Setting them equal in the sum formulas collapses each one into a single-angle statement, and those are the double angle identities of Lesson 10.4b, from which the power reduction and half angle formulas follow in turn. Nothing new will be proved there either.
Want this taught 1-on-1? Alexander tutors Trigonometry — $55/session, free consultation.