Three identity families obtained without any new proof. Setting the two angles equal in the sum identities gives the double angle formulas, including the three equivalent forms of the cosine one; solving those backwards gives the power reduction formulas that trade a square for a doubled angle; and substituting theta over two into those and taking square roots gives the half angle formulas, whose plus-or-minus is settled by the quadrant of the half angle rather than of the original.
Subject: Trigonometry · 70 slides · symbolic lesson
Open the interactive version of this deck
Title
Trigonometry · Section 10.4
Three families of identities, and not one new proof
Objectives
Lesson 10.4a proved one theorem about two angles. This lesson runs that theorem three times with the two angles chosen cleverly, and gets three whole families of identities for nothing.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 776-780 — Theorems 10.17, 10.18 and 10.19
Lesson 10.4a did the hard work: one computation with the distance formula gave the cosine of a difference, and everything else in that lesson followed from it. This lesson spends the result. You will not prove a new theorem anywhere in it.
The trick each time is to choose the two angles in the sum formula cleverly. Choose them equal and you get the double angle formulas. Read one of those backwards and you get the power reduction formulas. Feed half of an angle into those and take a square root and you get the half angle formulas.
That last square root is the only place where anything is lost, and it is the source of the one real difficulty in the lesson: a sign that the algebra cannot supply and you have to decide for yourself. Watch for it; nearly every mark lost in this section is lost there.
Warm-up
Discussion prompt
Write down, from memory, the sum formulas for cosine and for sine. Then say which one has a sign that flips, and which two functions are multiplied together in each term.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 775 — Theorem 10.16
Write both formulas out before you look. If you can only get one of them, get the cosine one, because the cosine one is the one that was proved from scratch and the others were built on it.
The detail worth rehearsing is the sign. In the cosine formula a plus inside the bracket becomes a minus on the right, which feels backwards the first twenty times you use it. In the sine formula the sign goes straight through.
The other detail is which functions get multiplied. Cosine keeps like with like: cosine times cosine, then sine times sine. Sine mixes them: sine times cosine, then cosine times sine. Every formula in today's lesson is one of these two with the same angle written twice.
Concept
Figure (svg): A vertical chain of four boxes joined by downward arrows: sum and difference identities, then double angle identities, then power reduction formulas, then half angle formulas, with the move that produces each one written beside its arrow.
No new picture is drawn in this lesson and no new proof is given. Each family is the one above it with a particular substitution made and, at the last step, a square root taken.
That is worth knowing under exam pressure. If a formula will not come back to you, rebuild it: the chain is three short moves long.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 776-779 — Theorems 10.17 to 10.19
Look at the picture as a map of the lesson rather than as content to learn. There are four boxes and three arrows, and the three arrows are the only new things here: set the two angles equal, solve for the square, substitute half of theta and take roots.
This matters practically. Under exam pressure formulas go missing, and the ones most likely to go missing are the half angle ones, because they are at the bottom of the chain and have the most moving parts. If you can remember the cosine sum formula and the three arrows, you can rebuild the bottom box in about ninety seconds.
It also tells you what to expect from the difficulty. Nothing below the first box is conceptually harder than what you already did; it is all bookkeeping, signs and choices of which form to use.
Section
Part 1
Concept
The sum formula holds for every pair of angles. Nothing stops you from choosing the same angle twice.
\[ \cos(\alpha+\beta) = \cos(\alpha)\cos(\beta) - \sin(\alpha)\sin(\beta) \]
\[ \alpha = \beta = \theta \]
\[ \cos(\theta+\theta) = \cos(\theta)\cos(\theta) - \sin(\theta)\sin(\theta) \]
\[ \cos(2\theta) = \cos^2(\theta) - \sin^2(\theta) \]
That is the whole derivation. The left side is a cosine of a doubled angle; the right side never mentions anything but the original angle.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 776 — Theorem 10.17
The move here is so small that people distrust it. The sum formula was proved for all applicable angles alpha and beta, and alpha and beta are just names. Nothing in the proof required them to be different, so you are allowed to point both names at the same angle.
Once you do, the left side becomes the cosine of theta plus theta, which is the cosine of twice theta, and the right side becomes a difference of two squares. Read that right side carefully: it is a squared cosine minus a squared sine, in that order.
Check the order with a value you know. At zero the squared cosine is one and the squared sine is zero, so the right side is one, and the cosine of zero is indeed one. If you ever write the subtraction the other way round, this one-second test will catch it.
Concept
\[ \sin(\theta+\theta) = \sin(\theta)\cos(\theta) + \cos(\theta)\sin(\theta) \]
\[ \sin(2\theta) = 2\sin(\theta)\cos(\theta) \]
The two mixed terms of the sine formula are the same product written twice, so they add rather than cancel.
\[ \tan(\theta+\theta) = \frac{\tan(\theta)+\tan(\theta)}{1-\tan(\theta)\tan(\theta)} \]
\[ \tan(2\theta) = \frac{2\tan(\theta)}{1-\tan^2(\theta)} \]
The tangent formula inherits a restriction: it says nothing at any angle where one of the three tangents in it is undefined.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 776 — Theorem 10.17
Do the same thing to the sine formula and the two mixed terms turn out to be the same product written twice, so instead of cancelling they add. That is where the factor of two comes from, and it is worth knowing that it is not a doubling of anything: it is two identical terms meeting.
The tangent formula comes the same way, and it inherits the restriction that comes with tangents. It is silent at any angle where the tangent of theta, or the tangent of twice theta, is undefined, and it is also silent where its own denominator vanishes.
That denominator is zero exactly when the tangent of theta is one or minus one, which happens at an eighth of a turn and at three eighths of a turn. You will see those two angles again in a graph a few slides from now.
Notation
Annotate
On: \( \cos(2\theta) = \cos^2(\theta) - \sin^2(\theta), \quad \sin(2\theta) = 2\sin(\theta)\cos(\theta) \)
Work through the annotations one at a time, and after each one look back at the formula. The single most useful observation is the first: the doubled angle appears only on the left, so the whole point of the theorem is that you never have to know anything about twice theta in order to compute with it.
The second annotation is the sign check, and you should make it a reflex. Any identity you half remember can be tested at zero, or at a quarter turn, in a couple of seconds. That is much faster than trying to recall which way round you saw it written.
The last annotation is about notation and it causes real confusion. A squared cosine means the cosine is taken first and then the result is squared. It does not mean the angle is squared, and it is not the same as the cosine of a squared angle.
Concept
The Pythagorean identity lets you swap a squared sine for a squared cosine at any time. Doing that to either term of the first form gives two more forms.
\[ \cos(2\theta) = \cos^2(\theta) - \sin^2(\theta) \]
\[ \sin^2(\theta) = 1 - \cos^2(\theta) \;\Longrightarrow\; \cos(2\theta) = \cos^2(\theta) - \left(1-\cos^2(\theta)\right) = 2\cos^2(\theta)-1 \]
\[ \cos^2(\theta) = 1 - \sin^2(\theta) \;\Longrightarrow\; \cos(2\theta) = \left(1-\sin^2(\theta)\right) - \sin^2(\theta) = 1-2\sin^2(\theta) \]
Three forms, one value. The first needs both functions; the second needs only the cosine; the third needs only the sine. That is why the cosine of a doubled angle can always be found from a single piece of information.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 776 — the three forms and the remark after them
Do not memorise three formulas. Memorise one, and remember that you can always swap a squared sine for one minus a squared cosine, or the other way round, at any moment.
Watch what each swap buys. Trading the squared sine away leaves twice a squared cosine minus one, which needs only the cosine. Trading the squared cosine away leaves one minus twice a squared sine, which needs only the sine. So whichever single value you happen to know, one of the three forms is ready for it.
This is the reason the book remarks that the cosine of a doubled angle needs only one piece of information while the sine appears to need two. That remark is the whole design of the next two parts of the lesson.
Picture it
Figure (svg): Two curves over one turn: the sine of two theta, a wave completing two full cycles between zero and two pi with height one, and twice the sine of theta, a single cycle of height two. At thirty degrees the two curves sit at different heights, 0.866 and 1.
The solid curve completes two cycles in the space the dashed one takes for one, and it never rises above one. The dashed curve rises to two. They agree only where both happen to be zero.
Two curves, and the difference between them is the difference between two ideas people constantly confuse. The solid curve is the sine of the doubled angle: it runs twice as fast and its height never passes one, because it is still a sine.
The dashed curve is twice the sine: it runs at the ordinary speed and reaches a height of two, which no sine ever does. The two curves cross only where both happen to be zero, which is precisely why the mistake survives so long. Test it at zero and it looks fine.
The marked pair at thirty degrees is the test that does catch it. There the sine of sixty degrees is about 0.866 and twice the sine of thirty degrees is exactly one. Different numbers, so different functions.
Trap
The commonest wrong first line in this whole section:
\[ \sin(2\theta) = 2\sin(\theta) \]
\[ \cos(2\theta) = 2\cos(\theta) \]
Wrong. A function name is not a multiplier.
Test it once, at thirty degrees, and you will never write it again.
\[ \sin(60^\circ) = 0.866 \]
\[ 2\sin(30^\circ) = 1 \]
The correct lines keep the original angle inside and multiply two different values together.
\[ \sin(2\theta) = 2\sin(\theta)\cos(\theta) \]
This is the most common error in the section, and it comes from reading the two inside the brackets as a multiplier that can be moved outside. It cannot, because sine is the name of a function and not a quantity being multiplied.
The cure is one worked test that you do once and remember. At thirty degrees the two sides come out as 0.866 and 1, which are visibly different. Do it now, on paper, and the error will stop happening.
Notice also what the correct formula looks like: two different functions multiplied together, both evaluated at the original angle. Whenever a doubled angle appears on one side of your work and only one function appears on the other, something has gone wrong.
Worked example
The point below lies on the terminal side of theta in standard position. Find the cosine and the sine of twice theta, and say where twice theta lands.
\[ P(-3, 4) \]
Find the distance from the origin
Why: Theorem 10.3 from Section 10.2.
\[ r = \sqrt{(-3)^2+4^2} = \sqrt{25} = 5 \]
Read off the two values
Why: Each coordinate over the radius.
\[ \cos(\theta) = -\frac35, \qquad \sin(\theta) = \frac45 \]
Use the first form of the cosine formula
Why: Both values are known, so the difference of squares is immediate.
\[ \cos(2\theta) = \left(-\tfrac35\right)^2-\left(\tfrac45\right)^2 = \frac{9}{25}-\frac{16}{25} = -\frac{7}{25} \]
Use the sine formula
Why: Twice the product of the two values.
\[ \sin(2\theta) = 2\left(\frac45\right)\left(-\frac35\right) = -\frac{24}{25} \]
Place the doubled angle
Why: Both coordinates of its point are negative.
\[ \left(-\tfrac{7}{25},\, -\tfrac{24}{25}\right) \;\Longrightarrow\; \text{Quadrant III} \]
Figure (svg): A unit circle with a ray into the second quadrant at about 127 degrees reaching the point minus three fifths, four fifths, and a second ray at about 254 degrees reaching the point minus seven twenty-fifths, minus twenty-four twenty-fifths, deep in the third quadrant.
Check with the tangent formula
Why: The tangent of theta is four over minus three, and the doubled tangent must agree with the quotient of the two answers.
\[ \frac{2\left(-\tfrac43\right)}{1-\tfrac{16}{9}} = \frac{24}{7} = \frac{-24/25}{-7/25} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 776-777 — Example 10.4.3 no. 1
This is the standard shape of a double angle problem: you are handed something that is not a value of a circular function, you turn it into values, and only then do you double. The distance to the origin is what turns coordinates into values.
Once the cosine and sine are in hand, the arithmetic is nothing. The part worth attention is the last question: where does the doubled angle land. You can answer it from the two signs alone, without knowing a single angle measure, because a point with both coordinates negative is in the third quadrant.
Look at the picture before you accept the answer. Theta points up and to the left, and twice theta has swung past a half turn down into the third quadrant. The picture also makes it obvious that the doubled angle is nowhere near twice as far up the circle in any intuitive sense; it has gone right round the top.
The check uses the third formula as an independent route to the same pair of answers. Getting the same number two ways is the cheapest confidence you will ever buy.
Prediction
\[ \theta \text{ in Quadrant II} \;\Longrightarrow\; 2\theta \text{ in which quadrant?} \]
Predict first
Theta lies in Quadrant II. Which quadrant contains twice theta?
Correct: Quadrant III or IV, depending on theta
Why: An angle between 90 and 180 degrees doubles to something between 180 and 360 degrees, so the doubled angle is below the horizontal axis, but which half depends on whether theta was under or over 135 degrees. The example had theta near 127 degrees, so twice it was near 254, in Quadrant III; had theta been 160 degrees, twice it would be 320, in Quadrant IV.
Commit to one of the four before revealing. Two of them are traps for a real habit of mind: the wish for a rule that does not depend on the particular angle.
Think about it as an interval rather than as a single angle. An angle between ninety and one hundred and eighty degrees doubles to something between one hundred and eighty and three hundred and sixty, which covers two whole quadrants. Nothing narrows it further without knowing more about theta.
The dividing line is one hundred and thirty five degrees. Below it the doubled angle is in the third quadrant, above it in the fourth. The worked example had theta near one hundred and twenty seven degrees, just under the line, which is why it landed in the third.
Matching
Match the pairs
Why: The three forms of the cosine formula are not decoration: each one is built for a different starting point, and choosing the right one means you never have to rebuild a value you were not given. The tangent formula plays the same role when the tangent is all you have.
Match each one before checking. The exercise looks trivial, and then in an exam under time pressure people reach for the first form out of habit and spend three lines rebuilding a value they did not need.
The rule to take away is one sentence: choose the form that names what you were given. If you were given only the sine, use the form with only the sine in it. That way there is never a second square root to take and never a second sign to decide.
The tangent row is slightly different because the tangent formula gives you the tangent of the doubled angle rather than its cosine. To get the cosine from a tangent you either use the tangent formula and then a triangle, or rebuild the sine and the cosine first, as the next example does.
Worked example
One number and one interval. Find the sine, cosine and tangent of twice theta.
\[ \tan(\theta) = -2, \qquad \frac{\pi}{2} < \theta < \pi \]
Find the squared secant
Why: The Pythagorean identity in its tangent-secant form.
\[ \sec^2(\theta) = 1+\tan^2(\theta) = 1+4 = 5 \]
Take the root with the quadrant's sign
Why: In Quadrant II the cosine is negative, so the secant is negative.
\[ \sec(\theta) = -\sqrt5 \;\Longrightarrow\; \cos(\theta) = -\frac{\sqrt5}{5} \]
Recover the sine
Why: Sine is tangent times cosine.
\[ \sin(\theta) = (-2)\left(-\frac{\sqrt5}{5}\right) = \frac{2\sqrt5}{5} \]
Double the sine
Why: Twice the product.
\[ \sin(2\theta) = 2\left(\frac{2\sqrt5}{5}\right)\left(-\frac{\sqrt5}{5}\right) = -\frac45 \]
Double the cosine with the sine-only form
Why: The squared sine is four fifths.
\[ \cos(2\theta) = 1-2\left(\frac45\right) = -\frac35 \]
Check with the tangent double angle formula
Why: The formula and the quotient of the two answers must agree.
\[ \frac{2(-2)}{1-4} = \frac43 = \frac{-4/5}{-3/5} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 784 — Exercise 58
This is the workhorse exercise of the section, and the section's exercise set has ten of them. The pattern never changes: convert the given value into a cosine and a sine, fixing the signs from the interval, and then double.
The conversion step is where the thinking is. One plus a squared tangent is a squared secant, which gives the size of the secant immediately; the interval gives its sign. Notice that the Pythagorean identity can only ever tell you the size of something, never the sign.
Once the cosine and the sine are on the page the rest is arithmetic, and the check with the tangent formula confirms both answers at once. If the quotient of your two answers does not match what the tangent formula gives, at least one of them is wrong.
Fill the middle
Fill in the blanks
\cos(2\theta) = \cos^2(\theta)-\sin^2(\theta) = \left(1-\sin^2(\theta)\right)-\sin^2(\theta) = 1-2\sin^2(\theta)
Why: The squared cosine is traded for one minus the squared sine. Once that substitution is made the two squared sines combine, leaving one minus twice a squared sine — the form to reach for when the sine is the only value you have.
Fill both blanks before checking. Both answers are the same number, which is the point: the only thing being used is that a squared cosine and a squared sine add to one.
The line in the middle is the one that is usually skipped when people write this out, and skipping it is how the sign gets lost. Write the substitution in full, with the bracket, and the two squared sines combine without drama.
When you have it, look at the shape of the result. There is a one, a subtraction, a two and a squared sine. If your remembered version has any of those four pieces in a different place, this derivation is thirty seconds long and will settle it.
Intuition
Figure (svg): An isosceles triangle with its apex at the left, two sides of length one meeting at an angle of two theta, and a vertical base on the right. A horizontal altitude of length cosine theta splits the apex angle into two equal parts and the base into two halves of length sine theta.
Draw a triangle with two sides of length one meeting at an angle of twice theta. Its area can be written two ways: with the two unit sides and the included angle, or with the base and the height that the altitude cuts out.
Setting the two areas equal gives the sine double angle formula with no algebra in it at all. The altitude bisects the apex angle because the triangle is isosceles, which is where the halving comes from.
This picture is not a proof you have to reproduce, but it is a good answer to the question of why the sine formula looks the way it does. The triangle has two sides of length one meeting at the doubled angle.
Compute its area with the two sides and the included angle and you get half of the sine of the doubled angle. Compute the same area as half the base times the height and you get the sine times the cosine, because the altitude cuts an isosceles triangle into two right triangles with hypotenuse one.
Two expressions for one area, so they are equal, and doubling both sides gives the formula. Notice that this also explains the factor of two geometrically: it is there because the base is made of two halves and you doubled to clear the halves.
Tweak it
Parameter explorer
The curve is twice the sine of x times the cosine of k times x. Slide k. At which value does the ragged product collapse into a single clean wave, and what wave is it?
\[ y = 2\sin(x)\cos(kx), \qquad k = {k} \]
Put the slider at one first and look at the shape: a clean wave of constant height, which is the sine of twice x. Then move the slider away from one in either direction.
What appears is a beat: a fast wiggle inside a slowly changing envelope, with the height no longer constant. That is what a product of two circular functions normally looks like, and it is the reason the product to sum formulas of Lesson 10.4c exist.
The lesson to take from this is that the double angle formula is a coincidence of matching angles, not a general fact about products. Collapse only happens when the two angles are the same, which is exactly the substitution that produced the formula.
Section
Part 2
Concept
Read the three cosine forms again and notice what each one costs. Any single value settles the cosine of the doubled angle.
| you are given | use | cost |
|---|---|---|
| the cosine only | two times a squared cosine, minus one | nothing extra |
| the sine only | one minus two times a squared sine | nothing extra |
| the tangent only | the tangent formula, then a quotient | nothing extra |
\[ \sin(2\theta) = 2\sin(\theta)\cos(\theta) \]
The sine of a doubled angle looks worse, because that formula names both functions. The next two examples show two different ways of getting round it.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 776 — the remark after Theorem 10.17
Read the table as a shopping list. Whatever you were handed, one of the three cosine forms is already in the right currency, and the cost column says you never have to buy anything else.
The sine formula is the odd one out because it names both functions at once. In practice this is usually not a problem, because once you know one value and a quadrant you can always get the other from the Pythagorean identity. But there are two ways of avoiding even that.
The first is to keep the answer in terms of a letter, which is Example 10.4.3 number 2 and the shape of every calculus substitution problem you will meet later. The second is to use the identity of Example 10.4.3 number 3, which gets the sine of a doubled angle out of a tangent alone.
Worked example
Here the letter x is not a coordinate. It is a name for the value of the sine.
\[ \sin(\theta) = x, \qquad -\frac{\pi}{2} \le \theta \le \frac{\pi}{2} \]
Write down what is wanted and substitute what is known
Why: The sine is x; the cosine is still missing.
\[ \sin(2\theta) = 2\sin(\theta)\cos(\theta) = 2x\cos(\theta) \]
Get the cosine from the Pythagorean identity
Why: Two square roots, for now.
\[ \cos^2(\theta) + x^2 = 1 \;\Longrightarrow\; \cos(\theta) = \pm\sqrt{1-x^2} \]
Kill the minus sign with the stated interval
Why: On that closed interval the terminal side is on or to the right of the vertical axis, so the cosine is never negative.
\[ -\frac{\pi}{2} \le \theta \le \frac{\pi}{2} \;\Longrightarrow\; \cos(\theta) \ge 0 \]
Substitute
Why: One expression, in x alone.
\[ \sin(2\theta) = 2x\sqrt{1-x^2} \]
Check at a value you already know
Why: Take x one half, so theta is a sixth of a half turn and twice theta is a third of one.
\[ 2\left(\tfrac12\right)\sqrt{1-\tfrac14} = \frac{\sqrt3}{2} = \sin\left(\frac{\pi}{3}\right) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 777 — Example 10.4.3 no. 2
The first thing the book says about this problem is worth repeating: the letter x here is not a coordinate. It is a name for the value of the sine, and the habit of reading x as a first coordinate has to be suspended for a moment.
The algebra is short. Substitute the sine, get the cosine from the Pythagorean identity, and you are left with a plus-or-minus that has to be resolved. Everything interesting is in resolving it.
The interval given runs from a quarter turn below the horizontal axis to a quarter turn above it, which is the right half of the circle. On the whole of that half the first coordinate is zero or positive, so the cosine is never negative and the plus root is the only survivor. The check at one half confirms the finished formula against an angle you already know.
Error analysis
Annotate
On: \( \sin(\theta)=x, \;; -\tfrac{\pi}{2}\le\theta\le\tfrac{\pi}{2} \;\Longrightarrow\; \sin(2\theta) = \pm 2x\sqrt{1-x^2} \)
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercises 86 and 87
See whether you can spot the problem before revealing the annotations. The algebra here is not wrong anywhere; it is simply unfinished, and unfinished in a way that looks like an answer.
A plus-or-minus in a final answer is a claim that the question does not determine which. Sometimes that is true, but here the question went out of its way to give you an interval, and an interval in a problem like this is there for exactly one purpose: to settle a sign.
Build the habit now. Every time a square root appears in your work, stop and ask which sentence of the question decides its sign. In this section that sentence is almost always an interval or a named quadrant, and if you cannot find one, the answer really does have two cases, as Exercise 94 will show.
Worked example
Verify the identity below. It gets the sine of a doubled angle out of one value.
\[ \sin(2\theta) = \frac{2\tan(\theta)}{1+\tan^2(\theta)} \]
Start from the right side and name the denominator
Why: One plus a squared tangent is a squared secant.
\[ \frac{2\tan(\theta)}{1+\tan^2(\theta)} = \frac{2\tan(\theta)}{\sec^2(\theta)} \]
Write both in sines and cosines
Why: Quotient and reciprocal identities.
\[ = \frac{2\left(\frac{\sin(\theta)}{\cos(\theta)}\right)}{\frac{1}{\cos^2(\theta)}} \]
Divide by a fraction by multiplying by its reciprocal
Why: The squared cosine comes up to the top.
\[ = 2\left(\frac{\sin(\theta)}{\cos(\theta)}\right)\cos^2(\theta) \]
Cancel one cosine
Why: And the right side is the double angle formula.
\[ = 2\sin(\theta)\cos(\theta) = \sin(2\theta) \]
Check at an eighth of a turn
Why: There the tangent is one and the doubled angle is a quarter turn.
\[ \frac{2(1)}{1+1} = 1 = \sin\left(\frac{\pi}{2}\right) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 777 — Example 10.4.3 no. 3
This is a verification rather than a computation, and the book starts from the right side because the right side is the messier one. That is the general rule for verifying: start with the complicated side and simplify, because simplifying is a directed activity and complicating is not.
The key recognition is in the first step. One plus a squared tangent is a squared secant, and recognising that instantly is worth practising, because the pattern turns up constantly. After that it is a matter of writing everything in sines and cosines and letting the cancellation happen.
The result earns its place in the lesson: it lets you get the sine of a doubled angle from the tangent alone, with no quadrant argument and no Pythagorean identity. It is also the identity the half angle tangent formula is derived from, later in this lesson.
Picture it
Figure (svg): Two curves on the same axes between minus and plus one and a half radians: the tangent of theta, rising once through the window, and the tangent of two theta, which rises three times and has vertical asymptotes at minus a quarter pi and a quarter pi as well as at the edges.
The tangent of theta passes through the window once; the tangent of twice theta passes through it three times. The red lines mark the two angles where the tangent of theta is one or minus one, which makes the denominator of the tangent double angle formula zero.
Count the branches. In this window the tangent of theta rises through once, and the tangent of twice theta rises through three times, because doubling the angle halves the spacing of the asymptotes.
The two red lines are at an eighth of a turn on either side of zero, where the tangent of theta is one and minus one. Those are exactly the angles that make the denominator of the tangent double angle formula vanish, and they are also where the doubled angle is a quarter turn, where the tangent is genuinely undefined.
So the formula's restriction is not a technicality invented by the authors. It matches a real feature of the function, and the picture is the proof.
Worked example
Use the identity just verified rather than rebuilding the sine and the cosine separately.
\[ \tan(\theta) = \frac{x}{7}, \qquad -\frac{\pi}{2} < \theta < \frac{\pi}{2} \]
Quote the identity
Why: It turns one tangent into the sine of the doubled angle.
\[ \sin(2\theta) = \frac{2\tan(\theta)}{1+\tan^2(\theta)} \]
Substitute
Why: Nothing else is needed; the interval already guarantees the tangent is defined.
\[ = \frac{2\left(\frac{x}{7}\right)}{1+\frac{x^2}{49}} \]
Clear the inner fractions
Why: Multiply the top and the bottom by 49.
\[ = \frac{14x}{49+x^2} \]
Check with a number
Why: Put x equal to 0.7, so the tangent is 0.1 and theta is 0.0997 radians.
\[ \frac{14(0.7)}{49.49} = 0.19802 = \sin(2\cdot 0.0997) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 91
This is the same problem as Example 10.4.3 number 2 with the tangent in place of the sine, and it is worth doing precisely because the identity from the previous slide makes it short.
If you had not had that identity, you would have had to build a right triangle with legs x and seven, find the hypotenuse, read off the sine and the cosine, and then multiply. That route works and gives the same answer; it just takes four times as long.
The interval here does no sign work at all, unlike the earlier example. It is there only to guarantee that the tangent exists and that the angle is uniquely determined, which is the usual role of that particular interval in calculus.
Step zero
\[ \sin(\theta) = -\frac{7}{25}, \qquad \frac{3\pi}{2} < \theta < 2\pi \]
Discussion prompt
Exercise 49 gives you this and asks for six values. Before touching any formula, what three decisions have you already been handed, and which one value still has to be computed?
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 784 — Exercise 49
Write your answer before revealing. The instinct in an exercise that asks for six values is to start computing, and the cost of that instinct is usually one or two wrong signs out of six.
Everything except one number is decided by the interval, and deciding it takes three lines of writing that you never have to revisit. The quadrant of theta fixes the sign of the cosine. Halving the interval fixes the quadrant of the half, and therefore the sign of all three of its values.
Notice the halving in particular. A fourth-quadrant angle, named between three quarters of a turn and a full turn, halves to something between three eighths and a half of a turn, which is in the second quadrant. Nothing about that is guessable from the phrase fourth quadrant on its own; you have to halve the interval.
Section
Part 3
Concept
The double angle formulas handle a factor of two. For three theta or four theta there is no new theorem: split the angle so that a sum formula applies, then let the double angle formulas finish the job.
\[ \cos(3\theta) = \cos(2\theta+\theta), \qquad \cos(4\theta) = \cos(2\cdot 2\theta) \]
Two splittings, two routes. Either way the answer is a polynomial: the cosine of any whole multiple of theta can be written using nothing but the cosine of theta.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 777 — Example 10.4.3 no. 4
There is no triple angle theorem, and you should not look for one. What there is instead is a routine: write the multiple angle as a sum of a smaller multiple and one more theta, expand with a sum formula, and keep going until nothing but theta is left.
For an even multiple there is a shortcut. Four theta is twice two theta, so you can double twice instead of expanding a sum at all, and the next few slides do it both ways so you can compare.
The result of the routine is always a polynomial, and that fact has a name outside this course: these are the Chebyshev polynomials, and they turn up in approximation theory and in numerical analysis. Exercise 96 asks you to look for the pattern.
Worked example
Express the cosine of three theta as a polynomial in the cosine of theta.
Split the angle as two plus one
Why: Then apply the cosine sum formula.
\[ \cos(3\theta) = \cos(2\theta)\cos(\theta) - \sin(2\theta)\sin(\theta) \]
Substitute the double angle formulas
Why: Choose the cosine form written in cosines only.
\[ = \left(2\cos^2(\theta)-1\right)\cos(\theta) - \left(2\sin(\theta)\cos(\theta)\right)\sin(\theta) \]
Multiply out
Why: A squared sine survives, so the job is not finished.
\[ = 2\cos^3(\theta) - \cos(\theta) - 2\sin^2(\theta)\cos(\theta) \]
Trade the squared sine for cosines
Why: The Pythagorean identity, one last time.
\[ = 2\cos^3(\theta) - \cos(\theta) - 2\left(1-\cos^2(\theta)\right)\cos(\theta) \]
Collect
Why: Two cubic terms and two linear ones.
\[ = 4\cos^3(\theta) - 3\cos(\theta) \]
Check at twenty degrees
Why: The cosine of twenty degrees is 0.939693, and three times twenty is sixty.
\[ 4(0.939693)^3 - 3(0.939693) = 0.5 = \cos(60^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 777 — Example 10.4.3 no. 4
Follow the choices, not just the algebra. The first choice is how to split the angle, and two plus one is the only useful split. The second choice is which form of the cosine double angle formula to substitute, and here you want the one written in cosines only, because the target is a polynomial in the cosine.
The third step is where people stop too early. Multiplying out leaves a term with a squared sine in it, and the answer is not a polynomial in the cosine until that term is traded away. The Pythagorean identity does it in one substitution.
The final answer is short enough to be worth recognising on sight. Four cubes minus three, applied to the cosine, gives the cosine of three times the angle. The numerical check at twenty degrees is quick and catches any sign slip in the collection step.
Picture it
Figure (svg): Two panels. On the left, the cosine of three theta plotted over zero to pi, with four marked points. On the right, the cubic four u cubed minus three u plotted over minus one to one, with the same four heights marked at u equal to one, one half, minus one half and minus one.
The left panel is a cosine wave running three times as fast as usual. The right panel is an ordinary cubic. They carry the same four marked heights, because feeding the cosine of theta into the cubic reproduces the wave exactly.
The two panels are the same information in two languages. On the left the horizontal axis is an angle and the curve oscillates. On the right the horizontal axis is the cosine of that angle and the curve is a plain cubic with no oscillation in it at all.
Trace one of the marked points across. At a third of a half turn on the left, the cosine of three theta is minus one. On the right, the cosine of that angle is one half, and the cubic evaluated at one half is also minus one. The identity is exactly the statement that this always works.
The reason the left curve oscillates and the right one does not is that the cosine itself does all the oscillating. Feed an oscillating input into a fixed cubic and the output oscillates; the cubic is only a fixed rule.
Ranking
Put in order
Put the steps of the expansion of the cosine of three theta into the order they must be done.
Why: The trade has to come after the substitution, because the squared sine only appears once the double angle formulas have been put in. Doing it earlier gives nothing to trade, and doing it later leaves a sine in the answer.
Drag the steps into order before checking. The two that are genuinely easy to swap are the substitution of the double angle formulas and the trade of the squared sine.
Doing the trade first accomplishes nothing, because before the substitution there is no squared sine on the page: there is only a sine of two theta, which is a different animal. The squared sine appears only when the sine double angle formula is put in and then multiplied by the other sine.
The last step, collecting, is not decoration. Two of the four terms after the trade are multiples of the plain cosine, and combining them is what turns a messy expression into the clean answer you are supposed to recognise.
Worked example
The same route, with the sine as the target function.
Split and apply the sine sum formula
Why: Sine mixes the two functions.
\[ \sin(3\theta) = \sin(2\theta)\cos(\theta) + \cos(2\theta)\sin(\theta) \]
Substitute, taking the cosine form in sines only
Why: The target is a polynomial in the sine, so choose the form that has no cosine in it.
\[ = 2\sin(\theta)\cos^2(\theta) + \left(1-2\sin^2(\theta)\right)\sin(\theta) \]
Trade the squared cosine for sines
Why: Every surviving cosine appears to an even power, so it can all be traded.
\[ = 2\sin(\theta)\left(1-\sin^2(\theta)\right) + \sin(\theta) - 2\sin^3(\theta) \]
Collect
Why: Two cubic terms and two linear ones again.
\[ = 3\sin(\theta) - 4\sin^3(\theta) \]
Check at twenty degrees
Why: The sine of twenty degrees is 0.342020.
\[ 3(0.342020) - 4(0.342020)^3 = 0.866025 = \sin(60^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 784 — Exercise 65
The route is identical to the cosine case, with one choice made differently. Because the target is a polynomial in the sine, the form of the cosine double angle formula to substitute is the one written in sines only.
After that substitution the only cosine left is squared, and an even power of the cosine can always be traded for sines. That is the reason this works at all, and it is the point that the counterexample slide will press on.
Compare the two answers. Four cubes minus three for the cosine, three minus four cubes for the sine. The near symmetry is pleasant but it is a coincidence of small numbers rather than a general pattern, so do not rely on it for four theta.
Worked example
Four theta is twice two theta, so no sum formula is needed at all.
Double, treating two theta as the angle
Why: Use the cosine-only form.
\[ \cos(4\theta) = 2\cos^2(2\theta) - 1 \]
Substitute the double angle formula inside
Why: Now the inner angle is theta.
\[ = 2\left(2\cos^2(\theta)-1\right)^2 - 1 \]
Expand the square
Why: The square of a binomial, term by term.
\[ = 2\left(4\cos^4(\theta) - 4\cos^2(\theta) + 1\right) - 1 \]
Collect
Why: A quartic in the cosine, with no odd powers.
\[ = 8\cos^4(\theta) - 8\cos^2(\theta) + 1 \]
Check at twenty degrees
Why: Four times twenty is eighty.
\[ 8(0.939693)^4 - 8(0.939693)^2 + 1 = 0.173648 = \cos(80^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 784 — Exercise 69
No sum formula appears here at all, which is the point of including it. Four theta is twice two theta, so applying the cosine double angle formula with two theta playing the part of the angle gets you most of the way in one line.
Watch the substitution carefully. In the first line the angle inside is two theta; in the second line it is theta. Writing both lines out, rather than doing them in your head, is what stops the two from getting confused.
The answer has only even powers in it, which had to happen: the cosine of four theta is unchanged when theta is replaced by the negative of theta, and only even powers have that property. That is a useful sanity check on any expansion of an even multiple.
Trap
Halfway through the cosine of three theta:
\[ \small 2\cos^3(\theta)-\cos(\theta)-2\sin^2(\theta)\cos(\theta) \]
Not finished. The question asked for a polynomial in the cosine, and a sine is still on the page.
The trade is the whole point of the exercise, and it is the step that is worth marks. One substitution finishes it.
\[ \sin^2(\theta) = 1-\cos^2(\theta) \]
\[ 4\cos^3(\theta)-3\cos(\theta) \]
Before you stop, read the question again and check that the only function left on the page is the one it named.
This is not an arithmetic error; it is an answer to a different question. The expression on the left is perfectly correct as an identity. It simply is not a polynomial in the cosine, which is what was asked for.
The habit that protects you is small: before writing a final answer, reread the question and check that the only function on your page is the one it named. That check takes five seconds and it catches this every time.
The same trap appears in reverse in the power reduction part of the lesson, where the question asks for first powers only and it is tempting to stop while a square is still sitting there.
Counterexample
\[ \sin(4\theta) = 4\sin(\theta)\cos^3(\theta) - 4\sin^3(\theta)\cos(\theta) \]
Discussion prompt
The sine of three theta came out as a polynomial in the sine of theta. Try the same for the sine of four theta, and say exactly what goes wrong. Then say what is special about the cases that do work.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercises 96 and 97
Try it before revealing, and try it seriously: expand the sine of four theta as twice the sine of two theta times the cosine of two theta and see what you are left with.
Every term ends up with an odd power of the cosine, and that is fatal. The Pythagorean identity trades squares for squares, so it can remove even powers of the cosine but never an odd one. A single leftover cosine is a square root of one minus a squared sine, and a square root is not a polynomial.
The general statement is worth remembering because it explains the whole pattern. The cosine of any whole multiple is a polynomial in the cosine. The sine of an odd multiple is a polynomial in the sine. The sine of an even multiple is not, and Exercise 97 is asking you to discover exactly that.
Section
Part 4
Concept
Both of the one-function forms contain a square. Reading them from right to left, each one hands you a formula for that square.
\[ \cos(2\theta) = 2\cos^2(\theta)-1 \;\Longrightarrow\; \cos^2(\theta) = \frac{1+\cos(2\theta)}{2} \]
\[ \cos(2\theta) = 1-2\sin^2(\theta) \;\Longrightarrow\; \sin^2(\theta) = \frac{1-\cos(2\theta)}{2} \]
power reduction — Trading a squared circular function for a first power of the same function at twice the angle.
Nothing has been proved that was not already known. The same equation is simply being used in the other direction.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 778 — Theorem 10.18
Nothing new is proved on this slide, and you should be suspicious if it feels like it. The two lines are the two one-function forms of the cosine double angle formula with the squared term moved to the left and everything else moved to the right.
The reason it is worth a name is that the direction matters enormously in practice. Going forwards you use these to turn a doubled angle into squares. Going backwards you use them to turn squares into a doubled angle, and it is the backwards direction that calculus needs constantly.
Why calculus needs it: a first power of a cosine has an obvious antiderivative and an obvious average, and a squared cosine has neither until you reduce it. That is the whole reason the reduction formulas have their own theorem number.
Notation
Annotate
On: \( \cos^2(\theta) = \frac{1+\cos(2\theta)}{2}, \qquad \sin^2(\theta) = \frac{1-\cos(2\theta)}{2} \)
The annotation to dwell on is the second one, about the angle doubling. It is the single most common source of error in this part of the lesson, and it is a mistake of reflex rather than of understanding.
The rule is: the angle in the reduced expression is twice the angle that was squared, whatever that angle happens to be. If you are reducing the square of a cosine of five theta, the answer has a cosine of ten theta in it.
The third annotation gives you a free check. Put theta equal to zero: the cosine version must give one and the sine version must give zero. That takes two seconds and tells you instantly whether you have the signs the right way round.
Intuition
Figure (svg): Three curves over one turn. The square of the cosine, a wave running between zero and one that completes two cycles; the cosine of two theta, running between minus one and one over the same two cycles; and a dashed horizontal line at one half, the midline of the squared curve.
Squaring a wave that runs between minus one and one gives a wave that runs between zero and one, and it does so twice as often, because the squaring folds the negative half up onto the positive half.
Reading the formula off the picture: halve the height of the double-angle wave, then lift it by one half so that its lowest point sits on zero. That is exactly one plus the cosine of twice theta, all over two.
Squaring does two visible things to a wave. It folds the part below the axis up above it, which doubles how often the pattern repeats, and it pulls the values toward zero, because squaring a number between zero and one makes it smaller.
The result is a wave running between zero and one, twice as fast as the original, which is precisely a cosine of twice the angle shrunk to half its height and lifted by one half. Read that description back and you have written the formula.
Notice the midline in particular. The dashed line at one half is the average value of a squared cosine over any whole number of cycles, and that single fact is what the polarising filter slide will turn into a physical prediction.
Worked example
Write the expression below as a sum and difference of cosines to the first power.
\[ \sin^2(\theta)\cos^2(\theta) \]
Replace each square by its reduction
Why: Two applications of Theorem 10.18.
\[ = \left(\frac{1-\cos(2\theta)}{2}\right)\left(\frac{1+\cos(2\theta)}{2}\right) \]
Multiply: a difference of squares
Why: The two mixed terms cancel.
\[ = \frac14\left(1-\cos^2(2\theta)\right) \]
A square is still there, so reduce again
Why: The inside angle is now two theta, so the reduction produces four theta.
\[ \cos^2(2\theta) = \frac{1+\cos(4\theta)}{2} \]
Substitute
Why: Distribute the quarter.
\[ = \frac14 - \frac14\cdot\frac{1+\cos(4\theta)}{2} = \frac14 - \frac18 - \frac18\cos(4\theta) \]
Simplify
Why: No powers left anywhere.
\[ \sin^2(\theta)\cos^2(\theta) = \frac18 - \frac18\cos(4\theta) \]
Check at twenty degrees
Why: Left side 0.103294; right side one eighth minus an eighth of the cosine of eighty degrees.
\[ (0.342020)^2(0.939693)^2 = 0.103294 = \tfrac18 - \tfrac18(0.173648) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 778 — Example 10.4.4
This is the model power reduction problem, and its shape is typical: reduce, simplify, discover that a square has reappeared, reduce again.
The second step is the pleasant one. The product of one minus something and one plus the same thing is a difference of squares, so the two mixed terms cancel and you are left with a single squared cosine to deal with.
The third step is the dangerous one, and it is the reason the error analysis slide exists. The angle inside that square is two theta, not theta, so reducing it produces a cosine of four theta. Writing the reduction out as its own line, as is done here, is what stops the slip.
Check the final answer at any convenient angle. At twenty degrees both sides come to 0.103294, which also tells you the answer is small there, as it should be for a product of a small squared sine with a squared cosine near one.
Error analysis
Annotate
On: \( \cos^2(2\theta) \;\overset{?}{=}\; \frac{1+\cos(2\theta)}{2} \)
Look at the line for a moment before revealing. It has the right shape, the right sign and the right denominator, and the only thing wrong with it is the angle, which is the easiest thing in the world to copy across unchanged.
The theorem is a statement about whatever is inside the square. When what is inside is two theta, the reduced form must carry four theta. Nothing in the theorem ever refers to a fixed letter theta; the letter is a placeholder.
The numerical disproof is worth doing yourself once. At twenty degrees the left side is 0.586824, the mistaken right side is 0.883022, and the correct right side, which uses eighty degrees, is 0.586824 again. One substitution of one number settles it.
Worked example
Verify the identity by reducing the left side all the way down.
\[ 8\sin^4(\theta) = \cos(4\theta) - 4\cos(2\theta) + 3 \]
Write the fourth power as the square of a square
Why: Then reduce the inner square.
\[ \sin^4(\theta) = \left(\frac{1-\cos(2\theta)}{2}\right)^2 \]
Expand the square
Why: A squared cosine of two theta appears.
\[ = \frac{1-2\cos(2\theta)+\cos^2(2\theta)}{4} \]
Reduce the remaining square
Why: Twice two theta is four theta.
\[ \cos^2(2\theta) = \frac{1+\cos(4\theta)}{2} \]
Substitute the reduced square
Why: Now every term is a first power, but the fraction is stacked.
\[ = \frac{1-2\cos(2\theta)+\frac{1+\cos(4\theta)}{2}}{4} \]
Clear the inner fraction
Why: Multiply the top and the bottom by two.
\[ = \frac{3 - 4\cos(2\theta) + \cos(4\theta)}{8} \]
Multiply by eight
Why: Which is the right side, rearranged.
\[ 8\sin^4(\theta) = \cos(4\theta) - 4\cos(2\theta) + 3 \]
Check at twenty degrees
Why: Eight times the fourth power of the sine of twenty degrees is 0.10947.
\[ 0.173648 - 4(0.766044) + 3 = 0.10947 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 784 — Exercise 63
A fourth power is a square of a square, and that is the whole strategy. Reduce the inner square, expand what you get, and then reduce the square that the expansion throws up.
The multiplication by eight at the end is not cosmetic. The natural result has a denominator of eight, and the exercise states the identity with the eight on the left precisely so that the answer has whole number coefficients, which is the form calculus actually wants.
Notice the three frequencies in the answer: a constant, a cosine of twice the angle and a cosine of four times the angle. That is a general pattern. Reducing an n-th power produces terms up to n times the angle, and nothing beyond.
Picture it
Figure (svg): Four curves over one turn: the fourth power of the sine, a run of narrow humps between zero and one; the constant three eighths; the two-term approximation three eighths minus a half of the cosine of two theta; and the small correction one eighth of the cosine of four theta.
The fourth power of the sine, divided by eight from the identity just proved, is a constant plus two ordinary cosine waves. The dashed curve is what you get from the constant and the first wave alone; the small third wave is the correction that makes it exact.
The thick curve is the fourth power of the sine. It is a train of humps, all above the axis, narrower than an ordinary sine wave because raising a number below one to the fourth power flattens everything except the peaks.
The dashed curve is what you get from the first two terms alone, the constant three eighths and the doubled-angle wave. It already has the right average and the right repeat length, and it is close, but its humps are not narrow enough.
The third curve is the correction that fixes them. Adding a small wave at four times the angle sharpens the peaks and flattens the troughs, and the sum matches the thick curve exactly. This is the first thing you ever see of Fourier analysis, and the product to sum formulas of the next lesson push it further.
Sorting
Sort into buckets
Each task wants the identities used in one direction. Which way?
Sort all six before checking. The sorting is not really about circular functions; it is about what the next step in the problem is going to be.
Anything that has to be integrated, averaged, or read as a combination of plain waves wants the powers gone, because a first power of a cosine behaves and a square does not. That is three of these six, and it is the direction calculus asks for far more often.
Anything that has to be expressed in terms of one angle wants the multiple angle broken down, which pushes the powers up. That is the direction of the polynomial expansions you did a few slides ago. The two directions are the same identities read from opposite ends.
Real world
Figure (svg): The square of the cosine of the filter angle plotted from zero to three hundred and sixty degrees, shaded beneath, running from one down to zero and back twice, with a dashed horizontal line at one half marking the average brightness.
Discussion prompt
Light passing through two polarising filters whose axes differ by an angle emerges with a brightness proportional to the square of the cosine of that angle. Rewrite this with power reduction, and say what it predicts about the average brightness as the second filter is turned steadily.
Write your rewrite before revealing. Power reduction turns the squared cosine into a constant plus a wave, and once it is in that form the physics reads straight off.
The constant is half the incoming brightness and the wave averages to zero, so the average light getting through is exactly half, no matter how fast you spin the filter. That is a real and slightly surprising prediction, and it is why a pair of polarising sunglasses cuts roughly half the light from an unpolarised source.
The doubled angle carries a second prediction. The light goes dark twice per full revolution of the filter, not once, because the brightness completes two cycles while the filter completes one. The two low points in the picture are those two dark positions.
Section
Part 5
Concept
The reduction formulas hold for every angle, so put half of theta in place of theta. The doubled angle on the right becomes theta itself.
\[ \cos^2\left(\frac{\theta}{2}\right) = \frac{1+\cos\left(2\cdot\frac{\theta}{2}\right)}{2} = \frac{1+\cos(\theta)}{2} \]
\[ \cos\left(\frac{\theta}{2}\right) = \pm\sqrt{\frac{1+\cos(\theta)}{2}} \]
Taking the square root is the only new move in the lesson, and it is the move that costs something: a square root cannot tell you which of two opposite numbers it came from, so a sign has to be supplied from outside.
\[ \sin\left(\frac{\theta}{2}\right) = \pm\sqrt{\frac{1-\cos(\theta)}{2}}, \qquad \tan\left(\frac{\theta}{2}\right) = \pm\sqrt{\frac{1-\cos(\theta)}{1+\cos(\theta)}} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 778-779 — Theorem 10.19
The substitution here is as small as the one that started the lesson: wherever the reduction formula says theta, write half of theta instead. On the right the doubled angle then becomes theta itself, which is exactly the trade you want.
Then comes the square root, and it is the only genuinely lossy move in the whole lesson. Squaring throws away a sign, so unsquaring cannot recover it, and the plus-or-minus in front of the root is the formula admitting that.
Read the plus-or-minus as an instruction rather than as part of the answer. It means: you decide, and then write the formula with one sign. An answer left with both signs in it is not an answer unless the question really does have two cases.
Notation
Annotate
On: \( \cos\left(\frac{\theta}{2}\right) = \pm\sqrt{\frac{1+\cos(\theta)}{2}} \)
The second annotation is the one worth reading twice, and then a third time before the next exercise. The sign comes from the quadrant of half of theta. Not from theta, not from the sign of the cosine of theta, not from anything else on the page.
The third annotation is a practical point that saves time. All three half angle formulas want the cosine of theta and nothing else, so if the problem gives you a sine or a tangent, your first job is to convert, using the interval to fix the sign.
The tangent version has an alternative with no square root at all, which you will derive in a few slides. When you have it, prefer it, because it removes the sign decision entirely: it produces the correct sign automatically.
Worked example
Lesson 10.4a found this with the difference formula. Now find it by halving a special angle.
Write the angle as a half
Why: Thirty degrees is one of the special angles.
\[ 15^\circ = \frac{30^\circ}{2} \]
Fix the sign from the half angle's quadrant
Why: Fifteen degrees is a Quadrant I angle, so its cosine is positive.
\[ \cos(15^\circ) = +\sqrt{\frac{1+\cos(30^\circ)}{2}} \]
Substitute the known cosine
Why: Root three over two.
\[ = \sqrt{\frac{1+\frac{\sqrt3}{2}}{2}} \]
Clear the inner fraction
Why: Multiply the top and bottom inside the root by two.
\[ = \sqrt{\frac{2+\sqrt3}{4}} = \frac{\sqrt{2+\sqrt3}}{2} \]
Check against the answer from the difference formula
Why: Exercise 95 asks you to prove the two forms equal; both come to 0.965926.
\[ \frac{\sqrt{2+\sqrt3}}{2} = 0.965926 = \frac{\sqrt6+\sqrt2}{4} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 779 — Example 10.4.5 no. 1, and Exercise 95
There are two routes to this value and you now know both. Lesson 10.4a wrote fifteen degrees as forty five minus thirty and used the difference formula; here it is thirty halved.
The sign is easy in this case, and it is worth noticing how easy, because the next example will not be. Fifteen degrees is in the first quadrant, where the cosine is positive, so the plus root is the one to take and no interval arithmetic is needed.
The two answers look nothing alike and they are equal, which Exercise 95 asks you to prove algebraically. Until you do, check them numerically: both come to 0.965926. Two different-looking exact forms for the same number is normal in this subject, and it is a good reason to always have a decimal value in reserve.
Intuition
Figure (svg): Four rungs of a ladder, evenly spaced from left to right: sixty degrees, thirty degrees, fifteen degrees and seven and a half degrees, with a curved arrow labelled halve joining each rung to the next one on its right, and the exact cosine of each angle written underneath it.
The special angles give you a handful of exact cosines. The half angle formula turns each of them into a new exact value, and then turns that one into another, for ever.
Nothing in the ladder is approximate. Each entry is a nest of square roots one level deeper than the one before it, which is why exact values at seven and a half degrees look so unpleasant and are still exact.
The special angles give you a small stock of exact cosines: thirty, forty five, sixty and a few others. The half angle formula turns that small stock into an infinite one.
Each rung costs one nested square root. The cosine of thirty degrees has one radical in it, fifteen has two, seven and a half has three, and the expressions get unreadable quickly. They remain exact, which is the point.
This ladder is not a curiosity. It is essentially how tables of sines were computed for centuries before calculators, by starting from a known angle and halving repeatedly until the steps were fine enough to interpolate between.
Matching
Match the pairs
Why: Halving compresses a whole turn into a half turn, so for any theta named between 0 and 360 degrees the half lands in the first or the second quadrant, where the sine is never negative. A negative half angle sine therefore requires theta to be described outside that range, as in Example 10.4.5 no. 2.
Match all four before checking, and then look at the four right-hand entries together, because as a set they say something the individual rows do not.
Halving compresses a whole turn into a half turn. So for any angle you name between zero and a full turn, the half lands somewhere in the top half of the circle, which means the first or second quadrant, which means its sine is never negative.
That has an immediate consequence. If a problem produces a negative half angle sine, as the next example does, then theta must have been described outside the range from zero to a full turn, for instance as a negative angle. Spotting that in advance tells you which sign to expect before you compute anything.
Intuition
Figure (svg): A unit circle with a ray into the third quadrant at about 258 degrees whose cosine is minus one fifth, together with two half-angle rays: one at about 129 degrees in the second quadrant where the height is positive, and one at about 309 degrees in the fourth quadrant where the height is negative.
The two rays drawn here are halves of the same angle. The first is half of the angle measured once around; the second is half of the same angle measured after one extra full turn.
Adding a whole turn to theta adds only a half turn to half of theta, which sends it to the opposite side of the circle and flips the sign of every one of its values. This is why the quadrant of theta by itself settles nothing.
Both rays in the picture are halves of the same angle, in the sense that doubling either one gets you back to the same terminal side. They are not the same angle, and their values are opposite.
The reason is arithmetic. Adding a full turn to theta adds only a half turn to half of theta, and a half turn sends every point to the other side of the circle, flipping the sign of both coordinates. So coterminal angles do not have coterminal halves.
This is why naming the quadrant of theta is not enough information for a half angle problem, and why every worked example in this part begins by writing down an interval for theta rather than a quadrant. Exercise 94 makes the point formally.
Worked example
Find the sine of half of theta from the information below.
\[ \cos(\theta) = -\frac35, \qquad -\pi \le \theta \le 0 \]
Halve the interval, in writing
Why: Divide every part of the inequality by two.
\[ -\frac{\pi}{2} \le \frac{\theta}{2} \le 0 \]
Read the sign off that interval
Why: Half of theta is in the fourth quadrant or on an axis, where the sine is never positive.
\[ \sin\left(\frac{\theta}{2}\right) \le 0 \]
Substitute into the half angle formula with the minus sign
Why: The cosine of theta is minus three fifths.
\[ \sin\left(\frac{\theta}{2}\right) = -\sqrt{\frac{1-\left(-\frac35\right)}{2}} \]
Simplify inside the root
Why: Eight fifths over two is four fifths.
\[ = -\sqrt{\frac{8/5}{2}} = -\sqrt{\frac45} \]
Rationalise
Why: Root four fifths is two over root five.
\[ = -\frac{2}{\sqrt5} = -\frac{2\sqrt5}{5} \]
Check numerically
Why: Theta is about minus 126.87 degrees, so half of it is about minus 63.43 degrees.
\[ \sin(-63.43^\circ) = -0.894427 = -\frac{2\sqrt5}{5} \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 779 — Example 10.4.5 no. 2
Notice the first line of the solution. Before touching a formula the book halves the interval, in writing, and only then reads the sign off. Copy that habit exactly; it is worth more marks than the algebra.
The interval given is a negative one, running from minus a half turn up to zero, which is the bottom half of the circle. Halved, it runs from minus a quarter turn to zero, the fourth quadrant, where the sine is negative. So the minus sign goes in front of the root before any numbers are substituted.
The simplification inside the root is the fiddly part: subtracting a negative three fifths gives eight fifths, and dividing by two gives four fifths. Rationalising at the end is conventional rather than necessary, but it is what the book's answer key expects.
The numerical check is the fastest possible confirmation. Theta is about minus one hundred and twenty seven degrees, so half of it is about minus sixty three, whose sine is indeed about minus 0.894.
Trap
A Quadrant III angle has a negative sine, so:
\[ \theta \text{ in QIII} \;\Rightarrow\; \sin\left(\tfrac{\theta}{2}\right) < 0 \]
Wrong. The sign belongs to the half, not to the whole.
Take theta equal to two hundred degrees, which is in Quadrant III. Half of it is one hundred degrees, which is in Quadrant II, where the sine is positive.
\[ \sin(100^\circ) = +0.9848 \]
Always write the halved interval down before choosing a sign. It takes one line and it is the line the marks are on.
This error feels like reasoning, which is what makes it dangerous. It looks as though you are using the quadrant information properly, and you are using it, just on the wrong angle.
The counterexample is quick and you should keep it. Two hundred degrees is comfortably in the third quadrant, where the sine is negative, and its half is one hundred degrees, where the sine is positive. One example is all it takes to kill the rule.
The protection is mechanical: never choose a sign until you have written the interval for half of theta on the page. If the problem gives you a quadrant rather than an interval, say so and give both cases, because that is what Exercise 94 shows the honest answer to be.
Worked example
An exact value from the half angle formula, with a sign to settle and a root to simplify.
Write it as a half and locate the half
Why: Half of 225 is 112.5, a Quadrant II angle, where the tangent is negative.
\[ 112.5^\circ = \frac{225^\circ}{2}, \qquad \tan(112.5^\circ) < 0 \]
Substitute the cosine of 225 degrees
Why: It is minus root two over two.
\[ \tan(112.5^\circ) = -\sqrt{\frac{1+\frac{\sqrt2}{2}}{1-\frac{\sqrt2}{2}}} = -\sqrt{\frac{2+\sqrt2}{2-\sqrt2}} \]
Rationalise inside the root
Why: Multiply the top and bottom by two plus root two.
\[ \frac{2+\sqrt2}{2-\sqrt2}\cdot\frac{2+\sqrt2}{2+\sqrt2} = \frac{6+4\sqrt2}{2} = 3+2\sqrt2 \]
Recognise the perfect square
Why: Three plus two root two is one plus root two, squared.
\[ -\sqrt{3+2\sqrt2} = -\sqrt{\left(1+\sqrt2\right)^2} = -\left(1+\sqrt2\right) \]
Check numerically
Why: One plus root two is 2.414214.
\[ \tan(112.5^\circ) = -2.414214 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 783 — Exercise 43
Three things happen here, and it is worth naming them separately: locating the half angle to fix a sign, substituting a known cosine, and simplifying a nested radical. Only the third is unfamiliar.
The simplification works because rationalising the fraction under the root produces three plus twice the root of two, and that happens to be a perfect square: it is one plus root two, all squared. Recognising a perfect square under a radical is a skill worth practising, because exact values in this subject produce them constantly.
If you do not spot the perfect square, the answer is still correct with the nested radical left in, and it is still exact. The numerical check tells you when you have simplified far enough, because both forms give 2.414214.
Worked example
Start from the identity of Example 10.4.3 no. 3 and turn it into a half angle formula, rather than transforming one side of the target into the other.
Replace theta by half of theta throughout that identity
Why: Every occurrence, on both sides.
\[ \sin\left(2\cdot\frac{\theta}{2}\right) = \frac{2\tan\left(\frac{\theta}{2}\right)}{1+\tan^2\left(\frac{\theta}{2}\right)} \]
Simplify the left side
Why: Two halves make one.
\[ \sin(\theta) = \frac{2\tan\left(\frac{\theta}{2}\right)}{1+\tan^2\left(\frac{\theta}{2}\right)} \]
Name the denominator and move it across
Why: One plus a squared tangent is a squared secant, whose reciprocal is a squared cosine.
\[ \sin(\theta) = 2\tan\left(\frac{\theta}{2}\right)\cos^2\left(\frac{\theta}{2}\right) \]
Power-reduce that squared cosine
Why: And the half disappears from the angle.
\[ \cos^2\left(\frac{\theta}{2}\right) = \frac{1+\cos(\theta)}{2} \]
Substitute, then divide
Why: The two and the halving cancel.
\[ \sin(\theta) = \tan\left(\frac{\theta}{2}\right)\left(1+\cos(\theta)\right) \]
Read off the formula
Why: No square root and no sign to choose.
\[ \tan\left(\frac{\theta}{2}\right) = \frac{\sin(\theta)}{1+\cos(\theta)} \]
Check against the previous example
Why: At theta equal to 225 degrees the sine is minus 0.707107 and one plus the cosine is 0.292893.
\[ \frac{-0.707107}{0.292893} = -2.414214 = \tan(112.5^\circ) \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 780 — Example 10.4.5 no. 3
This derivation runs in an unusual direction, and the book points that out. Instead of transforming one side of the target into the other, it starts from an identity already proved and manipulates it until the target appears.
The opening move is the one to remember: replace every theta by half of theta in an identity you already trust. That is always legal, because the identity was proved for all applicable angles, and it converts a fact about doubling into a fact about halving.
After that, each step removes one obstacle. The squared secant becomes a squared cosine, the squared cosine is power-reduced into one plus the cosine of theta, and dividing across finishes it.
The reward is large. The result has no square root, therefore no plus-or-minus, therefore no sign to decide. Whenever a problem wants the tangent of a half angle, this is the form to use.
Intuition
Figure (svg): A unit circle with a point P high in the second quadrant at one hundred and ten degrees, a chord drawn from the leftmost point of the circle to P, and a dashed right triangle under that chord whose horizontal leg is one plus the cosine of theta and whose vertical leg is the sine of theta; the chord meets the horizontal axis at half of theta.
Join the leftmost point of the unit circle to the point at angle theta. That chord subtends the same arc as the central angle does, so it rises at exactly half that angle.
Its rise is the sine of theta and its run is one plus the cosine of theta, so its slope is the radical-free formula just derived. The picture also explains why no sign has to be chosen: a slope carries its own sign.
The picture is the geometry behind the formula you just derived algebraically. Join the leftmost point of the circle to the point at angle theta, and you have a chord.
The central angle from the positive horizontal axis to that point is theta. The angle the chord makes with the horizontal is exactly half of it, because an inscribed angle subtending an arc is half the central angle on the same arc. That is a theorem from geometry, and this is the one place in the chapter where it pays off.
Now read the slope of the chord off the dashed right triangle: it rises by the sine of theta over a run of one plus the cosine of theta. Slope equals tangent of the angle of inclination, and there is the formula, sign and all. A slope carries its own sign, which is exactly why this version needs no plus-or-minus.
Edge cases
\[ \cos(\theta) = -\frac15, \qquad \theta \text{ in Quadrant III} \]
Discussion prompt
Work out the sine of half of theta twice: once assuming theta lies between a half turn and three quarters of a turn, and once assuming it lies between one and a half turns and one and three quarter turns. Both are Quadrant III angles. What does the comparison prove?
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 785 — Exercise 94
Do both computations before revealing, and do them fully, because the point only lands if you see the same arithmetic produce two different signs.
Both intervals describe third quadrant angles with the same cosine, so nothing distinguishes them as far as theta is concerned. But one of them is a full turn further round, and halving turns that full turn into a half turn, which moves the half angle from the second quadrant to the fourth.
The conclusion is a genuine limitation, not a trick. Theta is in the third quadrant is not enough information to determine the sign of a half angle value. A problem that wants a single answer has to give you an interval, and if it only gives you a quadrant, the correct response is to give both cases and say why.
Worked example
One cosine and one interval, and both directions of the lesson at once.
\[ \cos(\theta) = \frac35, \qquad 0 < \theta < \frac{\pi}{2} \]
Find the sine
Why: Quadrant I, so the positive root.
\[ \sin(\theta) = \sqrt{1-\tfrac{9}{25}} = \tfrac45 \]
Double the sine
Why: Twice the product.
\[ \sin(2\theta) = 2\left(\tfrac45\right)\left(\tfrac35\right) = \tfrac{24}{25} \]
Double the cosine from the cosine alone
Why: The cosine-only form.
\[ \cos(2\theta) = 2\left(\tfrac{9}{25}\right)-1 = -\tfrac{7}{25} \]
Double the tangent as a quotient
Why: Which is negative, so twice theta is in Quadrant II.
\[ \tan(2\theta) = \frac{24/25}{-7/25} = -\frac{24}{7} \]
Locate the half angle before halving
Why: An eighth of a turn at most, so the first quadrant and every value positive.
\[ 0 < \frac{\theta}{2} < \frac{\pi}{4} \]
Halve the cosine
Why: Plus sign, from the interval just written.
\[ \cos\left(\tfrac{\theta}{2}\right) = \sqrt{\tfrac{1+3/5}{2}} = \sqrt{\tfrac45} = \tfrac{2\sqrt5}{5} \]
Halve the sine
Why: Same interval, same sign.
\[ \sin\left(\tfrac{\theta}{2}\right) = \sqrt{\tfrac{1-3/5}{2}} = \sqrt{\tfrac15} = \tfrac{\sqrt5}{5} \]
Halve the tangent with the radical-free form
Why: No sign to choose and no root to simplify.
\[ \tan\left(\tfrac{\theta}{2}\right) = \frac{4/5}{1+3/5} = \frac12 \]
Check that the halves rebuild the whole
Why: Doubling the half angle values must return the original sine and cosine.
\[ 2\left(\tfrac{\sqrt5}{5}\right)\left(\tfrac{2\sqrt5}{5}\right) = \tfrac45, \qquad 2\left(\tfrac{2\sqrt5}{5}\right)^2-1 = \tfrac35 \]
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, p. 784 — Exercise 53
This is the capstone exercise, and it exercises both directions of the lesson from a single starting value. Work through it with the slide covered if you can.
The doubling half is routine once the sine is found: the cosine-only form avoids needing the sine at all for the cosine of twice theta, and the tangent comes out as a quotient. Notice that the doubled angle lands in the second quadrant, which the negative cosine and positive sine both confirm.
The halving half begins, as always, with the interval. An angle in the first quadrant halves into the first eighth of a turn, so every half angle value is positive and no sign decision is delicate. The tangent is done with the radical-free form because it is simply faster.
The check at the end is the best one in the lesson. Take your two half angle values and double them; you should recover the sine and the cosine you started with. Any sign error anywhere in the halving will show up immediately.
Section
Part 6
Comparison
Comparison matrix
| family | angle you start with | angle in the answer | what happens to the power |
|---|---|---|---|
| double angle | θ | 2θ | squares appear |
| power reduction | θ | 2θ | squares disappear |
| half angle | θ | θ/2 | a square root appears |
| multiple angle as a polynomial | θ | 3θ or 4θ | cubes and fourth powers appear |
Fill in the blanks before checking. The table is a summary, but filling it in yourself is what turns four separate procedures into one picture.
The pattern running down the last column is the one to carry away. Angles and powers trade against each other: every time the angle grows the powers grow with it, and every time the powers come down the angle goes up. Nothing in this lesson changes both in the same direction.
The last row is the odd one out only in size. Expanding a triple or quadruple angle is the same trade taken two steps instead of one, which is why the powers reach a cube or a fourth power rather than a square.
Pattern
Figure (svg): A decision diagram: compare the angle in the answer with the angle you know, then branch four ways to the double angle formulas, to peeling one theta off with the sum formula, to the power reduction formulas, and to the half angle formulas with a sign chosen from the half angle's quadrant.
This is the order to work in whenever a problem in this section lands in front of you. The opening move is always the same: compare the angle in the answer with the angle you were handed.
If the answer's angle is twice, the only real decision is which form of the cosine formula to use, and that decision is made by looking at what you were given rather than at what looks familiar. If it is three or four times, split and expand and remember to trade the squares away at the end.
If a power has to disappear, reduce, and double the inside angle each time you do. If the answer's angle is half, write the halved interval before you write anything else. Those five lines cover every exercise in this part of the section.
Estimation
\[ 2\sin(15^\circ)\cos(15^\circ) \]
Predict first
Without a calculator, what is this equal to?
Correct: 1/2
Why: Twice a sine times a cosine of the same angle is the sine of the doubled angle, so this is the sine of 30 degrees, which is one half. The distractor 0.966 is the cosine of 15 degrees itself, which is what you get by forgetting that the pattern collapses.
\[ 2\sin(15^\circ)\cos(15^\circ) = \sin(30^\circ) = \frac12 \]
Commit to one of the four before revealing. The exercise is deliberately small because the skill it tests is recognition rather than computation.
Twice a sine times a cosine of the same angle is a pattern, and the moment you see it you should stop and write the sine of the doubled angle instead. Here that gives the sine of thirty degrees, which is one half exactly.
The distractor to watch is 0.966, which is the cosine of fifteen degrees on its own. It is what you get if you start computing the two factors separately and lose track of what you were computing. The whole value of knowing these identities is that they let you not do that arithmetic.
Real world
Figure (svg): Two panels. On the left, three projectile paths launched at twenty, forty-five and seventy degrees; the twenty and seventy degree paths land at the same spot while the forty-five degree path goes furthest. On the right, the range plotted against launch angle, a single hump peaking at forty-five degrees, with equal heights at twenty and seventy degrees.
Discussion prompt
An object launched at speed v and angle alpha travels a horizontal distance equal to v squared times the sine of twice alpha, divided by the acceleration of gravity. Explain why 20 degrees and 70 degrees give the same range, and state the general rule about which pairs of launch angles match.
Answer before revealing, and answer with a reason rather than by computing both ranges. The reason is what transfers.
The range depends on the launch angle only through the sine of twice that angle. Two angles give the same range exactly when their doubles have the same sine, and two angles in a half turn have the same sine exactly when they add to a half turn. Doubling turns that into the rule: the two launch angles add to ninety degrees.
The left panel shows what the rule looks like physically: the low flat shot and the high lobbed shot land in the same place, spending very different amounts of time in the air. The right panel shows it as symmetry of a single hump about forty five degrees, the one launch angle whose partner is itself, and therefore the one that gives the greatest range.
Check
Check your understanding
You know only that the sine of theta is 0.6. Which line gets you the cosine of twice theta in one step?
Answer: C
Why: Only the third form names the sine and nothing else, so it gives 1 minus 2 times 0.36, which is 0.28, with no extra work. Choosing the form that matches what you were given is the whole point of there being three of them.
The question is about efficiency, not correctness. All three cosine forms are true, and any of them would eventually give the right answer; only one of them gives it without extra work.
With only the sine in hand, the form written in sines alone lands the answer in one substitution. Reaching for either of the others forces you to find the cosine first, which needs a quadrant you were not given, so the answer would have two cases where the question has one.
Check
Check your understanding
Theta is an angle between 180 and 270 degrees. Which quadrant contains half of theta?
Answer: B
Why: Halving the interval gives an angle between 90 and 135 degrees, which is Quadrant II. The sine and the cosecant of the half are therefore positive and everything else about it is negative.
Halve the interval on paper before choosing. That is the entire method, and it is the same method that every half angle problem in this lesson used.
One hundred and eighty to two hundred and seventy halves to ninety to one hundred and thirty five, which is the second quadrant. The tempting wrong answer is the third, copied straight from theta, and that is precisely the trap the trap slide was about.
Check
Check your understanding
Which of these is the square of the cosine of three theta, written with first powers only?
Answer: A
Why: The reduction doubles whatever angle is inside the square, and here that angle is three theta, so the answer carries six theta. The plus sign belongs to the cosine version.
Read the angle inside the square before anything else. It is three theta, so the reduced form must carry six theta, and only one option does.
The other three options each represent a specific slip: leaving the angle alone, taking the sine version's minus sign, or halving instead of doubling. If you picked one of them, name which slip it was, because that is the one to watch for in your own work.
Explain it to yourself
Discussion prompt
In two or three sentences, explain why the half angle formulas carry a plus-or-minus while the double angle and power reduction formulas do not, and say where the sign is actually decided.
Write your two sentences before revealing. Being able to say this cleanly means you will never again wonder whether some other formula in the lesson also needs a plus-or-minus.
The point is about information. The double angle and power reduction formulas are chains of equalities with nothing discarded along the way, so they determine their answers completely. The half angle formulas take a square root, and squaring is the one operation in the chain that destroys information, namely a sign.
The destroyed information has to be restored from outside the algebra, and the only thing that knows it is where half of theta actually lies. That is why the halved interval deserves its own line at the top of every half angle solution.
Exit ticket
\[ \cos(\theta) = -\frac45, \qquad \pi < \theta < \frac{3\pi}{2} \]
Discussion prompt
Find the sine of twice theta and the cosine of half of theta. Show the interval you used to fix each sign.
Do the whole thing on paper before revealing, and make sure both signs are justified in writing rather than in your head.
For the doubling, theta is in the third quadrant, so the sine is negative. Two negatives multiply to a positive, and the sine of twice theta comes out positive, which matches the doubled angle landing in the first quadrant, between a full turn and a full turn and a quarter.
For the halving, halve the interval first: from a half turn to three quarters becomes from a quarter to three eighths of a turn, which is the second quadrant, where the cosine is negative. Two negatives in one problem, arrived at by two completely different arguments, and mixing them up is the single likeliest way to lose this question.
Recap
| family | the formula | what to watch |
|---|---|---|
| double angle | cos 2θ in three forms; sin 2θ = 2 sin θ cos θ; tan 2θ = 2 tan θ over 1 − tan²θ | pick the cosine form that matches the value you have |
| multiple angles | cos 3θ = 4cos³θ − 3cos θ; cos 4θ = 8cos⁴θ − 8cos²θ + 1 | trade every squared sine away before you stop |
| power reduction | cos²θ = (1 + cos 2θ)/2; sin²θ = (1 − cos 2θ)/2 | the angle inside doubles each time you reduce |
| half angle | cos, sin and tan of θ/2 as plus-or-minus a root | the sign comes from the quadrant of the half |
\[ \tan\left(\frac{\theta}{2}\right) = \frac{\sin(\theta)}{1+\cos(\theta)} \]
Next, Lesson 10.4c adds and subtracts the sum formulas instead of specialising them, which turns products into sums and sums into products, the form Section 10.7 needs for solving equations.
Stitz & Zeager, College Trigonometry, Ch. 10 Foundations of Trigonometry — §10.4 Trigonometric Identities §10.4, pp. 776-780 — Theorems 10.17, 10.18 and 10.19
OpenStax Algebra and Trigonometry 2e, §9.3 Double-Angle, Half-Angle, and Reduction Formulas §9.3 — the same three families in OpenStax
Three families and one chain. Double angle comes from setting the two angles equal in a sum formula; power reduction is the cosine version read backwards; half angle is power reduction with half of theta substituted and a square root taken.
Two things to keep at your fingertips. First, the cosine of a doubled angle needs only one value, provided you pick the form that names it. Second, the plus-or-minus in a half angle formula is settled by the quadrant of the half, which means writing the halved interval down before you substitute anything.
Lesson 10.4c takes the sum formulas in a different direction: instead of setting the two angles equal, it adds and subtracts two of them so the mixed terms cancel. That turns products into sums and sums into products, and a product is what Section 10.7 needs, because only a product can be set to zero one factor at a time.
Want this taught 1-on-1? Alexander tutors Trigonometry — $55/session, free consultation.