Intuition The one core idea
To solve an exponential equation you must reach a number that is hiding in the exponent — the "top of the ladder". Every symbol on this page exists to build the one tool that fetches it down: the logarithm , the exact undo-button for raising a base to a power.
Before you can solve 2 x = 7 , you must be fluent in seven small pieces of notation. The parent note fires them off quickly. Here we slow all the way down and earn each one, in an order where every piece rests on the one before it.
Everything below happens on one machine : pick a fixed number, called the base , and multiply it by itself some number of times. That "number of times" is the exponent — and it is what we usually don't know.
Definition Base and exponent
In the expression b x :
b is the base — the number being multiplied by itself (the fixed ingredient).
x is the exponent (or power , or index ) — how many times the base appears in the product. It sits up and to the right , off the ground line.
Picture: think of a ladder. The base b stands on the floor; the exponent x is a label way up at the top rung. In an exponential equation, x is the label we can't reach by ordinary arithmetic.
b x for whole numbers
b x means "multiply b by itself x times":
b 3 = b × b × b , 2 4 = 2 × 2 × 2 × 2 = 16.
Picture: a stack of x identical blocks, each block worth "one multiplication by b ".
Why the topic needs this. If you can't read 2 x as "2 multiplied by itself x times", the sentence "x is stuck in the exponent" means nothing. This is rung one of the ladder.
Worked example Reading powers out loud
5 2 → "five to the power 2" → 5 × 5 = 25 .
3 0 = 1 → "any base to the power 0 is 1" (multiply b by itself zero times, and by convention you're left with the number 1 — the starting point before any multiplying happens).
But here is the catch that makes the whole topic necessary : in 2 x = 7 , the answer is not a whole number (2 2 = 4 is too small, 2 3 = 8 is too big). So x lands between 2 and 3. We must extend the meaning of b x to fractions and any real number .
b x for fractional and real exponents (base b > 0 )
The definition grows in two steps, each forced by an index law we already trust:
Fractions. b 1/ n is defined as the n -th root of b : b 1/2 = b , because ( b 1/2 ) 2 = b ( 1/2 ) ⋅ 2 = b 1 = b must hold. Then b p / q = ( b 1/ q ) p .
All real numbers. For an irrational exponent like x = 2.807 … , we squeeze b x between b raised to the fraction just below and just above it, and take the value they close in on.
Picture: plot the dots b 0 , b 1 , b 2 , … then fill in the smooth curve through them — the height at x = 2.807 is read straight off that unbroken curve.
Why we need b > 0 : negative bases break this (e.g. ( − 1 ) 1/2 = − 1 is not a real number), so exponential equations always use a positive base .
2 3 means 2 × 3 "
Why it feels right: the two numbers sit next to each other, so it looks like multiplication.
The truth: the small raised number counts copies of the base in a product , not a factor. 2 3 = 8 , but 2 × 3 = 6 .
Fix: whenever you see a raised number, say the word "power" and picture a stack of blocks (or, for non-whole powers, a point on the smooth curve).
The parent note quietly assumes you can flatten stacked powers. Two laws do all the work, and they hold for any real exponents (this is exactly why the extension in §1 was designed the way it was).
Why the topic needs these. The power-of-a-power law is the single fact used to derive the power law of logarithms (parent note: "( b y ) k = m k "). The same-base law is the shortcut when both sides share a base.
= and "do the same thing to both sides"
A = B says A and B are the same number wearing two different disguises.
Picture: a balance scale, perfectly level. If both pans hold the same weight, you may add the same amount, multiply by the same amount, or square both pans — and it stays level.
Later (once we have met the log in §5) we will add one more legal move to this list: applying the same function — such as log b — to both sides. It is the same principle: one operation, done identically to both pans, never tips the balance.
Why the topic needs this. "Take the log of both sides" is legal only because of this principle. You are applying one function to a single number written two ways — the balance never tips.
Intuition The heart of the whole method
Some operations come in undo-pairs . Adding 3 is undone by subtracting 3. Squaring is undone by square-rooting. Raising the base b to a power is undone by... the ==logarithm base b ==. That last pair is the entire reason logs exist.
Definition Inverse function
An inverse f − 1 takes the output of a function f and hands you back the original input . The two defining properties are:
f ( f − 1 ( x ) ) = x and f − 1 ( f ( x ) ) = x .
In words: do then undo, or undo then do — you land exactly where you started.
Picture: a machine and its exact rewind. Feed x into the "b -to-the-power" machine → out comes b x . Feed b x into the log machine → out comes x again, untouched. For us f ( x ) = b x and f − 1 ( m ) = log b m , so log b ( b x ) = x and b l o g b m = m .
See Inverse Functions for the full treatment. Here is the one pairing we need, pictured:
Common mistake "Square root and log are the same kind of undo"
Why it feels right: both "undo" something.
The truth: x undoes squaring, where the unknown is the base (x 2 ). A log undoes exponentiation, where the unknown is the exponent (b x ). Different position on the ladder → different tool.
Now the star of the show. It is nothing but the undo-button of Section 4, given a name and a notation.
log b — "which power?"
log b ( m ) asks a question : "To what power must I raise the base b to get m ?"
The answer is that power.
log b ( m ) = x ⟺ b x = m
Read the two-way arrow ⟺ as "these two statements say the exact same thing, just rearranged."
Definition Which bases and inputs are allowed (domain)
For log b ( m ) to make sense we require:
Base b > 0 and b = 1 . We need b > 0 so b x is defined for all real x (§1). We exclude b = 1 because 1 x = 1 for every x — the machine gives one output for all inputs, so it has no undo. We exclude b ≤ 0 for the same reason b > 0 was forced in §1.
Input m > 0 . Since b x is always positive (a positive base to any power stays above zero), the question "b to the what is m ?" only has an answer when m itself is positive. log b ( 0 ) and log b ( negative ) are undefined .
Picture: on the smooth curve of §1, every height b x sits above the ground line — so you can only "read back" an x from a height that is genuinely above zero.
Worked example Reading a log as a question
log 2 ( 8 ) asks "2 to the what is 8?" Answer: 3 , because 2 3 = 8 . So log 2 8 = 3 .
log 10 ( 1000 ) asks "10 to the what is 1000?" Answer: 3 .
log 5 ( 1 ) asks "5 to the what is 1?" Answer: 0 , because 5 0 = 1 .
Definition Shorthand names for two special bases
ln always means base e , the natural log , where e ≈ 2.718 . See The Number e and Natural Logarithms .
log with no base written is context-dependent. In school maths, on calculators, and in most exam boards it means base 10 : log = log 10 . But beware: in higher mathematics, physics, and much scientific software (e.g. many programming languages) "log " means the natural log log e . Always check the convention of the source you are reading.
Good news: for solving b x = c any base works and gives the same final x — the choice only changes the intermediate numbers, never the answer.
Why the topic needs this. Section 4 said "logs undo exponentials". Section 5 gives that undo a symbol you can write on paper and a question you can say out loud . Every step of the parent's method is this symbol at work.
log b ( m ) is m divided by b somehow"
Why it feels right: the base and the input sit close together like a fraction.
The truth: log b ( m ) is a height (an exponent), not a ratio. log 2 ( 8 ) = 3 is nowhere near 8/2 = 4 .
Fix: always translate to the question "b to the what is m ?"
Now that "power", "base", "exponent", index laws, and the log symbol all exist, the parent note's key formula can be derived cleanly — using nothing but pieces already on this page.
For the product and quotient partners of this law, see Laws of Logarithms ; for why log b log c = log b c , see Change of Base Formula .
≈
≈ means "approximately equal to " — the value has been rounded.
Picture: a number line where the true answer lands very near a tidy decimal, and we point to the tidy one.
Why the topic needs it: answers like x = log 2 log 7 ≈ 2.807 are irrational — they never round off cleanly, so we honestly write ≈ , not = .
Definition How much to round (a practical rule)
Keep more digits mid-working, round only at the end. Rounding early (e.g. calling log 2 ≈ 0.30 ) and then dividing lets the error grow.
Match the precision your problem asks for. Exam questions usually say "to 3 significant figures" or "to 3 decimal places" — obey that. If nothing is stated, 3 significant figures is a safe default.
Sanity-check the last digit: carry one extra digit than you report, then round. For log 7/ log 2 = 2.8073 … reporting to 3 d.p. gives 2.807 .
Node IDs are just labels; each is spelled out in the box. Read top-to-bottom — an arrow means "learn the top one first".
Powers b to the x as repeated multiply
Extend to real exponents smooth curve
Index laws power of a power
The exponent is the unknown height
Equals sign as a balance scale
Apply the same function to both sides
Inverse functions the undo pair
Logarithm the undo of b to the x
Domain rules base positive input positive
Universal method b to the x equals c
Solving exponential equations 3.2.10
Every arrow is a "you need this first". Trace any path top-to-bottom: nothing appears before it was built.
Test yourself — cover the right side.
In b x , what is the base and what is the exponent ? b is the base (multiplied by itself); x is the exponent, the number of times.
What does 2 4 equal, and why is it NOT 2 × 4 ? 2 × 2 × 2 × 2 = 16 ; the exponent counts copies in a product, not a factor.
How is b 1/2 defined, and why? b 1/2 = b , because
( b 1/2 ) 2 = b 1 = b must hold by the power-of-a-power law.
Why must the base b be positive for b x to work for all real x ? Negative bases give non-real values (e.g.
( − 1 ) 1/2 = − 1 ), so we require
b > 0 .
Simplify ( b y ) k using an index law. b y k — power of a power multiplies the exponents.
If b p = b q (same fixed base b ), what can you conclude? p = q ; equal outputs force equal exponents.
Why is "apply the same function to both sides" a legal move? An equation is a balance; one operation done identically to both pans keeps it level.
State the two defining properties of an inverse function f − 1 . f ( f − 1 ( x )) = x and f − 1 ( f ( x )) = x — do-then-undo lands you back at the start.
What question does log b ( m ) ask? "To what power must I raise b to get m ?"
What are the domain restrictions on log b ( m ) ? Base b > 0 and b = 1 ; input m > 0 (since b x is always positive).
Evaluate log 2 ( 8 ) and log 10 ( 1000 ) . 3 and 3 (since 2 3 = 8 and 1 0 3 = 1000 ).
What does log (no base) mean, and what warning comes with it? Usually base 10 in school/calculators, but base e in higher maths and much software — always check the convention.
Which operation is the inverse (undo) of b x ? log b — the logarithm to base b .
Give the algebraic derivation of log b ( m k ) = k log b m . m k = ( b l o g b m ) k = b k l o g b m ; take log b of both sides to get k log b m .
What does ≈ signal, and what is a safe default precision? "Approximately equal" (rounded); round only at the end, default 3 significant figures.