Make logarithms meaningful

Connect logarithms with inverse exponentials, graphs, equivalent forms and valid domains.

What to practice

  • Connect logarithms with inverse exponentials, graphs, equivalent forms and valid domains.

An exponential function answers a forward question: start with a base, apply it xx times, and what do you get? Many problems ask the reverse. How many doublings turn 1 gram into 1,000 grams? Which power of 10 is 31.6? A logarithm is the name for that unknown exponent. Because Chapter 10 showed how inverses reverse a mapping, you already know the logic: a logarithmic function is the inverse of an exponential function, and every property in this chapter is an exponent property seen from the other side.

Keep one sentence in mind: a logarithm is an exponent. When an expression looks unfamiliar, translate it into a statement about a power.

A logarithm names an exponent

Let b>0b>0 with b≠1b\neq 1, and let c>0c>0. Then

log⁡bc=ameansba=c\log_b c=a\quad\mathrm{means}\quad b^a=c

Read log⁡bc\log_b c as "the exponent you put on bb to get cc." The base is written as a subscript. Two bases have special notation. When no base is written, the base is 10, so log⁡1000=3\log 1000=3. The natural logarithm ln⁡c\ln c means log⁡ec\log_e c, where e≈2.718e\approx 2.718.

The restrictions are not arbitrary. Every power of a positive base is positive, so no exponent produces 0 or a negative number; log⁡2(−8)\log_2(-8) is undefined. A base of 1 is excluded because every power of 1 equals 1.

Some logarithms can be found exactly by arithmetic, when cc is a recognizable power of bb. Others can only be estimated between integers or approximated with technology.

Worked example 11.1: evaluate by matching powers

Evaluate without a calculator.

  1. log⁡927\log_9 27. Write both numbers as powers of 3. If 9a=279^a=27, then 32a=333^{2a}=3^3, so 2a=32a=3 and log⁡927=32\log_9 27=\frac{3}{2}. Check: 93/2=(9)3=279^{3/2}=(\sqrt{9})^3=27.
  2. log⁡218\log_2\frac{1}{8}. Since 2−3=182^{-3}=\frac{1}{8}, the value is −3-3. A logarithm can be negative; its argument cannot.
  3. log⁡0.001\log 0.001. The base is 10, and 10−3=0.00110^{-3}=0.001, so the value is −3-3.
  4. ln⁡e4\ln e^4. The exponent on ee is 4, so ln⁡e4=4\ln e^4=4.

The common wrong answer to the first part is 279=3\frac{27}{9}=3. That divides the numbers instead of finding an exponent, and 93=7299^3=729 shows it fails.

Worked example 11.2: estimate between integers

Estimate log⁡5300\log_5 300 and log⁡40.1\log_4 0.1.

  • log⁡5300\log_5 300. Bracket 300 between powers of 5: 53=1255^3=125 and 54=6255^4=625. Since 125<300<625125<300<625, the logarithm is between 3 and 4. To refine, find the exponent 3.5: 53.5=1255≈279.55^{3.5}=125\sqrt{5}\approx 279.5. Since 300 is larger, the logarithm is a little more than 3.5. A calculator gives log⁡5300≈3.544\log_5 300\approx 3.544.
  • log⁡40.1\log_4 0.1. Now the argument is less than 1, so the exponent is negative. Since 4−2=0.06254^{-2}=0.0625 and 4−1=0.254^{-1}=0.25, and 0.0625<0.1<0.250.0625<0.1<0.25, the logarithm is between −2-2 and −1-1. A calculator gives about −1.661-1.661.

Bracketing first protects you from calculator slips. If a calculator returned 1.661 for the second value, the missing sign would be obvious.

Read a logarithmic scale

On an ordinary number line, equal distances represent equal differences. On a logarithmic scale, equal distances represent equal factors. On a base-10 scale, the position kk is labeled 10k10^k, so moving one unit to the right multiplies the value by 10. The units of a logarithmic scale are multiplicative.

This makes logarithmic scales useful for values that range over several powers of 10. In Figure 11.1, the numbers 1, 10 and 100 are crowded together on a linear scale but evenly spaced on a logarithmic one.

Two horizontal number lines. The top line is a linear scale from 0 to 10,000 with tick labels every 2,000; the values 1, 10 and 100 appear as nearly overlapping dots just right of 0, with a note that they crowd together near 0, and separate dots mark 1,000 and 10,000. The bottom line is a base-10 logarithmic scale on which 1, 10, 100, 1,000 and 10,000 are equally spaced, with the powers 10 to the 0 through 10 to the 4 printed above them. An open diamond halfway between 10 and 100 is labeled halfway, 10 to the 1.5, approximately 31.6.

Figure 11.1. The same five numbers on a linear scale and on a base-10 logarithmic scale. Original diagram.

The point halfway between 10 and 100 on the logarithmic scale sits at position 1.5, so it represents 101.5≈31.610^{1.5}\approx 31.6, not the arithmetic midpoint 55. To read any point, find its position and raise the base to that power. A point three-tenths of the way from 100 to 1,000 is at position 2.3 and represents 102.3≈199.510^{2.3}\approx 199.5. Reading it as 100+0.3(900)=370100+0.3(900)=370 treats the scale as linear between the labels, which it is not.

Logarithms undo exponentials

For a base b>0b>0 with b≠1b\neq 1, the functions f(x)=bxf(x)=b^x and g(x)=log⁡bxg(x)=\log_b x are inverses:

log⁡b(bx)=x for every real x,blog⁡bx=x for x>0\log_b(b^x)=x\ \mathrm{for\ every\ real}\ x,\qquad b^{\log_b x}=x\ \mathrm{for}\ x>0

So 5log⁡57=75^{\log_5 7}=7 and log⁡3(3−2.4)=−2.4\log_3(3^{-2.4})=-2.4, with no calculation: each function undoes the other. All the inverse facts from Chapter 10 apply. If (s,t)(s,t) is on the graph of y=bxy=b^x, then (t,s)(t,s) is on the graph of y=log⁡bxy=\log_b x. The domain of log⁡bx\log_b x is x>0x>0, the range of bxb^x; the range of log⁡bx\log_b x is all real numbers, the domain of bxb^x. The graphs are reflections over y=xy=x, as Figure 11.2 shows for base 2.

Square grid from negative 3 to 9 on both axes. A solid curve shows y equals 2 to the x, rising from near the x-axis on the left. A dashed curve shows y equals log base 2 of x, rising from near the y-axis at the bottom. A dotted line shows y equals x. Four pairs of points, joined by thin gray segments, are marked: (0, 1) with (1, 0), (1, 2) with (2, 1), (3, 8) with (8, 3), and (negative 2, one-fourth) with (one-fourth, negative 2).

Figure 11.2. An exponential function and its inverse logarithmic function, with four exchanged pairs. Original graph.

The two functions turn change patterns inside out. For an exponential function, adding a constant to the input multiplies the output by a constant. For a logarithmic function, multiplying the input by a constant adds a constant to the output. On the dashed curve in Figure 11.2, each doubling of the input, from 1 to 2 to 4 to 8, raises the output by exactly 1.

A logarithmic function in general form is f(x)=alog⁡bxf(x)=a\log_b x, with a≠0a\neq 0, b>0b>0 and b≠1b\neq 1.

Worked example 11.3: build an inverse table for 4x4^x

Start with values of y=4xy=4^x, then exchange the rows.

xx−1012\frac{1}{2}12
4x4^x14\frac{1}{4}12416

The exchanged table gives log⁡414=−1\log_4\frac{1}{4}=-1, log⁡41=0\log_4 1=0, log⁡42=12\log_4 2=\frac{1}{2}, log⁡44=1\log_4 4=1 and log⁡416=2\log_4 16=2. The entry log⁡42=12\log_4 2=\frac{1}{2} is easy to miss until you read it as "4 to what power is 2?"

Worked example 11.4: recognize a logarithmic pattern in data

Constructed practice data: xx takes the values 1, 4, 16 and 64, and yy takes the values 5, 8, 11 and 14. Find a model.

  1. Test the inputs. Each input is 4 times the previous one, so the inputs change by a constant factor.
  2. Test the outputs. Each output is 3 more than the previous one, so the outputs change by a constant difference.
  3. Name the pattern. Multiplied inputs paired with added outputs is the logarithmic pattern.
  4. Build the rule. Each factor of 4 adds 3, and y=5y=5 when x=1x=1, where log⁡41=0\log_4 1=0. So y=5+3log⁡4xy=5+3\log_4 x.
  5. Check. At x=64x=64, log⁡464=3\log_4 64=3 and y=5+9=14y=5+9=14.

If the roles were reversed, with inputs rising by equal differences and outputs multiplied, the pattern would be exponential.

Read the graph of a logarithmic function

The graph of f(x)=alog⁡bxf(x)=a\log_b x has a fixed set of features.

  • Domain and range. The domain is x>0x>0, and the range is all real numbers.
  • Monotonicity and concavity. The function is always increasing or always decreasing, and its graph is always concave down or always concave up, so it has no point of inflection. The table shows which case applies.
  • Extrema. On its whole domain there is no maximum or minimum. On a closed interval inside the domain, the extremes occur at the endpoints.
  • Asymptote and end behavior. The line x=0x=0 is a vertical asymptote. Near it the outputs are unbounded; as xx increases without bound, the outputs are also unbounded, though they change slowly.
Sign of aaBaseBehaviorConcavity
a>0a>0b>1b>1increasingconcave down
a>0a>00<b<10<b<1decreasingconcave up
a<0a<0b>1b>1decreasingconcave up
a<0a<00<b<10<b<1increasingconcave down

Compare this table with the one for exponential functions in Chapter 8. Reflecting over y=xy=x keeps an increasing function increasing but flips its concavity, which is why 2x2^x is concave up and log⁡2x\log_2 x is concave down.

An additive transformation moves these features. For g(x)=alog⁡b(x−h)+kg(x)=a\log_b(x-h)+k, the domain is x>hx>h and the vertical asymptote is x=hx=h. The multiply-then-add pattern now belongs to x−hx-h, not to xx: proportional values of xx itself no longer give equal output changes.

Worked example 11.5: features of a shifted, reflected logarithm

Describe g(x)=2−log⁡2(x+3)g(x)=2-\log_2(x+3), shown in Figure 11.3.

  1. Domain and asymptote. The argument must be positive, so x+3>0x+3>0 and the domain is x>−3x>-3. The vertical asymptote is x=−3x=-3.
  2. Direction and concavity. log⁡2(x+3)\log_2(x+3) increases and is concave down. Multiplying it by −1-1 makes gg decreasing and concave up.
  3. End behavior. As xx approaches −3-3 from the right, log⁡2(x+3)\log_2(x+3) decreases without bound, so g(x)g(x) increases without bound. As xx increases, g(x)g(x) decreases without bound.
  4. Intercepts. Setting g(x)=0g(x)=0 gives log⁡2(x+3)=2\log_2(x+3)=2, so x+3=4x+3=4 and x=1x=1. The yy-intercept is g(0)=2−log⁡23≈0.415g(0)=2-\log_2 3\approx 0.415.
  5. Pattern check. The marked inputs make x+3x+3 equal to 0.5, 1, 2, 4, 8 and 16. Each doubling lowers gg by exactly 1.

Graph of a decreasing, concave-up curve labeled y equals 2 minus log base 2 of the quantity x plus 3, drawn to the right of a dashed vertical line labeled x equals negative 3. The curve rises steeply near the dashed line without touching it and falls slowly to the right. Marked points are (negative 2.5, 3), (negative 2, 2), (negative 1, 1), (1, 0), (5, negative 1) and (13, negative 2).

Figure 11.3. A reflected and shifted logarithmic function with its vertical asymptote. Original graph.

Because the domain is an open interval, gg has no maximum or minimum. Restricted to −1≤x≤5-1\leq x\leq 5, it has maximum g(−1)=1g(-1)=1 and minimum g(5)=−1g(5)=-1, both at endpoints.

Rewrite logarithmic expressions

For b>0b>0, b≠1b\neq 1 and positive inputs xx and yy, four properties hold.

PropertyStatement
Productlog⁡b(xy)=log⁡bx+log⁡by\log_b(xy)=\log_b x+\log_b y
Quotientlog⁡bxy=log⁡bx−log⁡by\log_b\frac{x}{y}=\log_b x-\log_b y
Powerlog⁡b(xn)=nlog⁡bx\log_b(x^n)=n\log_b x
Change of baselog⁡bx=log⁡axlog⁡ab\log_b x=\frac{\log_a x}{\log_a b} for any valid base aa

Each property is an exponent rule in disguise. Let m=log⁡bxm=\log_b x and n=log⁡byn=\log_b y, so x=bmx=b^m and y=bny=b^n. Then xy=bm+nxy=b^{m+n}, and the definition gives log⁡b(xy)=m+n\log_b(xy)=m+n.

Each property also describes a graph.

  • The product property says log⁡b(kx)=log⁡bk+log⁡bx\log_b(kx)=\log_b k+\log_b x. A horizontal dilation of a logarithmic graph is the same as a vertical translation.
  • The power property says raising the input to a power multiplies the output, so log⁡b(xk)\log_b(x^k) is a vertical dilation of log⁡bx\log_b x for x>0x>0.
  • Change of base says log⁡bx=1log⁡ablog⁡ax\log_b x=\frac{1}{\log_a b}\log_a x. Every logarithmic function is a vertical dilation of every other, which is also why a calculator with only log⁡\log and ln⁡\ln keys can evaluate any base.

Worked example 11.6: rewrite for a purpose

  1. Show that y=log⁡3(9x)y=\log_3(9x) is a vertical translation of y=log⁡3xy=\log_3 x. By the product property, log⁡3(9x)=log⁡39+log⁡3x=2+log⁡3x\log_3(9x)=\log_3 9+\log_3 x=2+\log_3 x. Compressing horizontally by a factor of 19\frac{1}{9} has the same effect as shifting up 2.
  2. Write log⁡9x\log_9 x as a multiple of log⁡3x\log_3 x. By change of base, log⁡9x=log⁡3xlog⁡39=12log⁡3x\log_9 x=\frac{\log_3 x}{\log_3 9}=\frac{1}{2}\log_3 x, a vertical dilation by a factor of 12\frac{1}{2}.
  3. Expand log⁡(1000x2)\log(1000x^2) for x>0x>0. The product and power properties give log⁡1000+2log⁡x=3+2log⁡x\log 1000+2\log x=3+2\log x.

Worked example 11.7: condense, then check the domain

Write 2ln⁡x+ln⁡(x+4)−ln⁡32\ln x+\ln(x+4)-\ln 3 as a single logarithm for x>0x>0.

  1. Power first. 2ln⁡x=ln⁡(x2)2\ln x=\ln(x^2).
  2. Combine. ln⁡(x2)+ln⁡(x+4)−ln⁡3=ln⁡x2(x+4)3\ln(x^2)+\ln(x+4)-\ln 3=\ln\frac{x^2(x+4)}{3}.
  3. Compare domains. The original requires x>0x>0. The condensed form only requires x2(x+4)>0x^2(x+4)>0, which also allows −4<x<0-4<x<0. The two expressions agree only for x>0x>0, so the restriction must be stated.

The same caution applies to ln⁡x+ln⁡(x−4)\ln x+\ln(x-4) and ln⁡(x(x−4))\ln(x(x-4)). The first is defined only for x>4x>4. The second is defined for x<0x<0 or x>4x>4. Rewriting with the properties never removes the original domain's restrictions.

Common errors to repair

  • Splitting a sum. log⁡b(x+y)\log_b(x+y) is not log⁡bx+log⁡by\log_b x+\log_b y. The product property needs a product.
  • Confusing a quotient of logarithms with the logarithm of a quotient. log⁡bxlog⁡by\frac{\log_b x}{\log_b y} is a change of base, log⁡yx\log_y x, not log⁡bxy\log_b\frac{x}{y}.
  • Squaring the logarithm. (log⁡x)2(\log x)^2 is not 2log⁡x2\log x. The power property moves the exponent of the argument.
  • Dropping the domain. ln⁡(x2)\ln(x^2) and 2ln⁡x2\ln x agree only for x>0x>0.
  • Reading a logarithmic scale linearly. Halfway between 10 and 100 is about 31.6, not 55.

An explained incorrect solution. A student writes: "The graph of y=log⁡2(8x)y=\log_2(8x) is the graph of y=log⁡2xy=\log_2 x stretched vertically by a factor of 8." The 8 multiplies the input, not the output, so it cannot be a vertical stretch. By the product property, log⁡2(8x)=3+log⁡2x\log_2(8x)=3+\log_2 x, a vertical shift up 3, equivalently a horizontal compression by a factor of 18\frac{1}{8}. A quick check exposes the error: at x=1x=1, the student's graph gives 8log⁡21=08\log_2 1=0, but log⁡28=3\log_2 8=3.

Try it

11.1 (guided). Evaluate log⁡279\log_{27}9. If 27a=927^a=9, then 33a3^{3a} equals 3 raised to the power ____, so a=a= ____.

11.2. Let h(x)=3log⁡5(x−2)+1h(x)=3\log_5(x-2)+1. Find h(7)h(7), state the domain and vertical asymptote, and find the exact xx-intercept.

11.3. For x>0x>0, write log⁡2(32x5)−3log⁡2(2x)\log_2(32x^5)-3\log_2(2x) in the form p+qlog⁡2xp+q\log_2 x.

11.4. A classmate writes, "log⁡20=log⁡(10+10)=log⁡10+log⁡10=2\log 20=\log(10+10)=\log 10+\log 10=2." Explain the error, give a valid rewrite, and show without a calculator that 1<log⁡20<21<\log 20<2.

Check your try-it work

11.1. 33a=323^{3a}=3^2, so 3a=23a=2 and log⁡279=23\log_{27}9=\frac{2}{3}. Check: 272/3=(273)2=927^{2/3}=(\sqrt[3]{27})^2=9.

11.2. h(7)=3log⁡55+1=4h(7)=3\log_5 5+1=4. The domain is x>2x>2, and the asymptote is x=2x=2. Setting h(x)=0h(x)=0 gives log⁡5(x−2)=−13\log_5(x-2)=-\frac{1}{3}, so x=2+5−1/3≈2.585x=2+5^{-1/3}\approx 2.585.

11.3. log⁡232+5log⁡2x−3(log⁡22+log⁡2x)=5+5log⁡2x−3−3log⁡2x=2+2log⁡2x\log_2 32+5\log_2 x-3(\log_2 2+\log_2 x)=5+5\log_2 x-3-3\log_2 x=2+2\log_2 x.

11.4. The product property applies to products, not sums. A valid rewrite is log⁡20=log⁡(2⋅10)=log⁡2+1\log 20=\log(2\cdot 10)=\log 2+1. Since 10<20<10010<20<100, 1<log⁡20<21<\log 20<2; a calculator gives about 1.301.

Retrieve and review

Without the book, write the definition of log⁡bc\log_b c with its restrictions, the four properties and the graph meaning of each. Sketch y=2xy=2^x and y=log⁡2xy=\log_2 x with two exchanged pairs. At your next session, redo Worked example 11.5 for 2+log⁡2(x−1)2+\log_2(x-1) and compare every feature. A week later, explain in two sentences why log⁡b(kx)\log_b(kx) is a vertical translation. Chapter 12 uses these properties to solve exponential and logarithmic equations and to read semi-log plots of data.

Practice

Attempt the printed Chapter 11 practice set before reading any solution. It contains fourteen multiple-choice questions, Questions 167–180, followed by Written tasks 11–12.

  • No calculator: Questions 167–177 and both written tasks. Give exact values and state every domain restriction you rely on.
  • Graphing calculator required: Questions 178–180. Report decimal answers to the precision each question names.

Allow about 30 minutes for the multiple-choice questions and about 16 minutes for each written task. Then check your work in "Chapter 11: answers and explanations" in the answer section. For each miss, use the repair note printed with the explanation and return to the named heading before trying a fresh question.

Use this chapter companion

Review the lesson, attempt the original practice, then check the explanation. Your printed study guide remains complete without the website.

Independence

Skill by Skill Prep is independent. SAT® and AP® are trademarks registered by the College Board, which is not affiliated with, and does not endorse, this website or book. PSAT/NMSQT® is a registered trademark of the College Board and the National Merit Scholarship Corporation, which are not affiliated with, and do not endorse, this site. Bluebook™ is a trademark of the College Board. ACT® is a trademark of ACT Education Corp., which is not affiliated with and does not endorse this website or book. Desmos is a trademark of Desmos Studio PBC. Practice results describe these questions, not official scores or admission predictions.

Last updated .