From Arithmetic to Algebra

Learning goals

  • Read algebra as arithmetic's move from particular numbers to general claims
  • State the familiar laws as claims about every number
  • Recognize an identity as an equation true for all values
  • Test equivalence by substitution, and say what testing cannot prove
  • Disprove a general claim with a single counterexample

Arithmetic asks “what is”; algebra asks “what is always”

In arithmetic you work with numbers you can see, and every question has a numerical answer. The whole subject is about computing: combine these particular numbers and report the result. Algebra keeps all of that arithmetic but adds a new move. Instead of asking what happens to 33 and 55, it asks what happens no matter which two numbers you pick.

The tool that makes this possible is the variable: a letter that stands for a number whose name we do not know yet. Sometimes the name is unknown because we are trying to find it, as when we solve x+3=7x + 3 = 7 for the one value that fits. Other times the name is unknown on purpose, because we want a single statement to hold true no matter which number replaces the letter.

Watch how the second use replaces an endless list. You could check that

2+3=3+2,6+1=1+6,40+9=9+40,2 + 3 = 3 + 2, \qquad 6 + 1 = 1 + 6, \qquad 40 + 9 = 9 + 40,

and keep going forever without ever finishing. Or you can write the whole pattern once:

a+b=b+a.a + b = b + a.

Here aa and bb do not stand for two particular numbers. Substitute any pair you like into a+b=b+aa + b = b + a, and it comes out true; that is what makes the line a claim about every pair, not a report about one. That shift, from computing with fixed numbers to reasoning about all numbers together, is one of the two moves that take you from arithmetic into algebra. Naming an unknown you mean to find, as with x+3=7x + 3 = 7, is the other.

The laws every number obeys

Throughout arithmetic you leaned on a small set of rules, usually without stating them. In algebra we write those rules with letters, which turns each one into a statement about every number. These are the laws that all of algebra is built on.

LawWith additionWith multiplication
Commutativea+b=b+aa + b = b + aab=baab = ba
Associative(a+b)+c=a+(b+c)(a + b) + c = a + (b + c)(ab)c=a(bc)(ab)c = a(bc)
Identitya+0=aa + 0 = aa⋅1=aa \cdot 1 = a
Inversea+(−a)=0a + (-a) = 0a⋅1a=1a \cdot \dfrac{1}{a} = 1 (for a≠0a \neq 0)

One more law links addition and multiplication, the distributive law:

a(b+c)=ab+ac.a(b + c) = ab + ac.

Read each row as a promise that holds no matter what numbers the letters stand for. The commutative law promises that order never changes a sum or a product. The associative law promises that regrouping never changes them either. The identity law names the two special numbers that leave things alone, 00 for addition and 11 for multiplication. The inverse law names the partner that undoes a number, −a-a for adding and 1a\tfrac{1}{a} for multiplying. These are not new facts to memorize; they are the arithmetic you already trust, written so that a letter can stand in for any number.

It is worth pausing to see that these laws are earned, not simply declared. Here is why one of them, the commutative law of addition, must hold for every pair of numbers.

Why a+b=b+aa + b = b + a for every pair of numbers#

Think first about counting. Adding two whole numbers means combining two collections and counting the total. Put a pile of aa stones together with a pile of bb stones. Whether you scoop the first pile into the second or the second into the first, you end up looking at the very same heap of stones. Counting that one heap gives one answer. The total cannot depend on which pile you named first, so a+ba + b and b+ab + a are the same count.

The same idea carries over to numbers that are not counts of stones, including negative ones, using the number line, though negatives need one more idea: a rightward move and a leftward move partly undo each other. When aa and bb point the same way, both positive or both negative, there is nothing to undo: moving aa then bb, or bb then aa, just lays two same-direction rods end to end, and exactly as with the stone piles the order does not change how far the combined rod reaches. When either one is 00, that move covers no ground at all, so it trivially does not matter where it sits in the order.

When one is positive and the other is negative, call the positive one’s size pp and the negative one’s size nn, so the negative number is −n-n. The two moves overlap over whichever is shorter, a stretch of length min⁡(p,n)\min(p, n), and that overlap always cancels, however the moves are ordered.

If p≥np \ge n: moving right pp then left nn retraces the last nn units of that rightward trip, canceling them, and leaves p−np - n to the right of the start. Moving left nn first then right pp: the first nn units of that rightward trip retrace the same ground back to the start, and the remaining p−np - n continues past it, landing at the identical p−np - n to the right.

If n>pn > p: moving right pp then left nn: the first pp units of that leftward trip retrace the rightward trip back to the start, and the remaining n−pn - p continues past it, landing at n−pn - p to the left. Moving left nn first then right pp retraces the last pp units of that leftward trip, leaving the identical n−pn - p to the left.

Either way, swapping the order only changes which trip gets retraced first; the two trips still cancel over their shared length and leave the same leftover in the same direction, so a+ba + b and b+ab + a mark the same point. (Check it: with p=3p = 3 and n=5n = 5, so n>pn > p, right 33 then left 55 lands at 3−5=−23 - 5 = -2, and left 55 then right 33 lands at −5+3=−2-5 + 3 = -2, the same point, n−p=2n - p = 2 to the left.)

Nothing in either argument depended on which numbers aa and bb actually were: the counting argument works the same way whatever the two piles hold, and the number-line argument works the same way whatever the two moves are. So the equality holds for every choice of aa and bb, not just the ones checked above. That is exactly what the single line a+b=b+aa + b = b + a records.

Identities: equations that are always true

An equation is two expressions joined by an equals sign, and once a variable appears, an equation can be true for no values, for one value, for several values, or for every value. Compare these three:

x+3=7,a+0=a,x+1=x.x + 3 = 7, \qquad a + 0 = a, \qquad x + 1 = x.

The first, x+3=7x + 3 = 7, is true for exactly one value, x=4x = 4, and false for every other. Finding that value is what the coming lessons call solving the equation. The second, a+0=aa + 0 = a, is true for every value of aa without exception. The third, x+1=xx + 1 = x, is true for no value at all, since no number equals one more than itself. (An equation can also land in between, true for some values and false for others, though none of the three examples above happen to do that.)

An equation that is true for every value of its variables is called an identity. All the laws in the table above are identities, and so is the distributive law: each holds for every number you could substitute. An identity makes no demand on the variable; it simply reports a fact that never fails. A plain equation like x+3=7x + 3 = 7 is a different animal, a question that singles out particular values, and telling the two apart is the first habit of algebra.

Check your understanding

Which of these equations is an identity (true for every value of the variable)?

Answer choices

Equivalent expressions

Two expressions are equivalent when they produce the same value for every input. Equivalent expressions are exactly the two sides of an identity. If AA and BB agree no matter what the variables are, then the equation A=BA = B is an identity, and the other way around.

The distributive law is a factory for equivalent expressions. Take 2(x+3)2(x + 3) and 2x+62x + 6. They look different, one a product and one a sum, yet they always agree, and the reason is worth seeing in full.

Why 2(x+3)2(x + 3) and 2x+62x + 6 are equivalent#

The expression 2(x+3)2(x + 3) means two copies of the quantity x+3x + 3 added together:

2(x+3)=(x+3)+(x+3).2(x + 3) = (x + 3) + (x + 3).

Now regroup, which the associative and commutative laws permit, collecting the two xx‘s together and the two 33‘s together:

(x+3)+(x+3)=(x+x)+(3+3)=2x+6.(x + 3) + (x + 3) = (x + x) + (3 + 3) = 2x + 6.

Every step used a law that holds for every number, and no step depended on what xx actually is. So 2(x+3)2(x + 3) and 2x+62x + 6 deliver the same value for every xx, which is precisely what it means for them to be equivalent.

The same picture appears in a rectangle, where the distributive law becomes a statement about area.

Area model for the distributive lawA rectangle of height a and total width b plus c is divided by a vertical line into a left rectangle labeled ab and a right rectangle labeled ac, so a times the quantity b plus c equals ab plus ac.abcabaca(b + c) = ab + ac
A rectangle of height a and width b + c has area a(b + c). A vertical cut at width b splits that area into ab and ac, so a(b + c) and ab + ac measure the same region. That is why they are equivalent for every a, b, and c.

Not every pair that looks close is equivalent. The expressions 2(x+3)2(x + 3) and 2x+32x + 3 differ only by where the parentheses reach, but they are not the same. Finding one value where the two expressions part ways is all it takes to prove they are not equivalent.

Worked example 1 Are these expressions equivalent?

Start with 4(n+1)4(n + 1) and 4n+44n + 4. The distributive law multiplies the 44 across both parts of the sum:

4(n+1)=4n+4.4(n + 1) = 4n + 4.

Because the distributive law holds for every number, this equality holds for every nn, so the two expressions are equivalent.

Now compare 4(n+1)4(n + 1) with 4n+14n + 1. Here the 44 has reached only the nn, not the 11. Test a single value, say n=2n = 2:

4(2+1)=12,4(2)+1=9.4(2 + 1) = 12, \qquad 4(2) + 1 = 9.

The two disagree at n=2n = 2, so 4(n+1)4(n + 1) and 4n+14n + 1 are not equivalent. One value where they differ is enough to settle it.

Substitution and the power of a single counterexample

To substitute is to replace a variable with a chosen number and then do the arithmetic. Writing the number in parentheses where the letter used to be keeps the operations and signs intact, so 5x5x at x=7x = 7 becomes 5(7)=355(7) = 35 rather than the meaningless ”5757.” Substitution is how you test whether a claimed identity is actually true.

The logic of that testing is not symmetric, and the asymmetry is one of the most useful ideas in all of mathematics. To disprove a claim that an equation holds for every number, you need only one value where the two sides disagree. To prove that it holds for every number, no amount of testing is ever enough.

Why one counterexample settles a claim but examples never prove it#

Suppose someone claims an equation is an identity, true for every number. The claim is a statement that holds no matter which number you substitute. A single value where the left side and the right side come out different directly contradicts the word “every,” so that one counterexample ends the matter: the claim is false, and nothing more needs checking.

The other direction fails for a simple reason. There are infinitely many numbers, so any list of values you actually test, however long, leaves infinitely many untested. Agreement on your list could be a coincidence that breaks at the very next number. Testing can therefore raise your confidence, but it can never rule out a hidden failure.

To be certain an equation holds for all numbers, you must derive it from the laws. A derivation gives that certainty because those laws are the statements already known to hold for all numbers. This is why algebra proves identities by manipulation, as in the earlier proof that 2(x+3)=2x+62(x + 3) = 2x + 6, rather than by piling up examples.

The next example shows exactly how a couple of lucky agreements can mislead you.

Worked example 2 Is n+n=n⋅nn + n = n \cdot n an identity?

A claim that two expressions are always equal is a claim about every number, so try a few values and stay alert for one that breaks it.

At n=2n = 2 the two sides agree:

2+2=4,2⋅2=4.2 + 2 = 4, \qquad 2 \cdot 2 = 4.

At n=0n = 0 they agree again, since 0+0=00 + 0 = 0 and 0⋅0=00 \cdot 0 = 0. It is tempting to declare the equation an identity, but two successes prove nothing. Try n=3n = 3:

3+3=6,3⋅3=9.3 + 3 = 6, \qquad 3 \cdot 3 = 9.

Now the sides differ, so n+n=n⋅nn + n = n \cdot n is false in general. Two agreements were never a guarantee that it holds everywhere, which is exactly why checking a value or two can fool you, and why the one disagreement at n=3n = 3 is what actually settles the question.

Check your understanding

A student claims that n2=4nn^2 = 4n for every number nn. Which value of nn shows the claim is false?

Answer choices

The order of operations does not change

Bringing in letters changes nothing about the order of operations. Parentheses still come first, then multiplication and division, then addition and subtraction, exactly as in arithmetic. So 2+3x2 + 3x means 2+(3x)2 + (3x), with the multiplication done before the addition, and it is a different expression from (2+3)x(2 + 3)x. When you substitute a number, the order of operations decides the result.

Worked example 3 Evaluate 7−2x7 - 2x when x=3x = 3

Substitute 33 for xx, in parentheses, then follow the order of operations:

7−2x=7−2(3).7 - 2x = 7 - 2(3).

Multiplication comes before subtraction, so multiply first:

7−2(3)=7−6=1.7 - 2(3) = 7 - 6 = 1.

The value is 11. It is not (7−2)(3)=15(7 - 2)(3) = 15; the subtraction does not happen first just because it is written first. The same order of operations you used in arithmetic still governs once a letter is present.

Check your understanding

Evaluate 3+4x3 + 4x when x=2x = 2.

Answer choices

Turning words into algebra

Most problems arrive as words, and turning those words into symbols is a skill worth having early. Name the unknown with a letter, then convert each phrase into an operation: “more than,” “the sum of,” or “increased by” adds; “times” or “the product of” multiplies; “divided by” or “per” divides. “Less than” also subtracts, but it reverses the order you read: ”55 less than nn” means n−5n - 5, not 5−n5 - n, because nn is the amount you start from and 55 is what comes off it. Two kinds of results come out of this, and they are not the same.

A word phrase turns into an expression: it names a value, and makes no claim about it. A word statement that says two amounts are equal, using a word like is or equals for the equals sign, turns into an equation: a claim you can test. An expression is something you evaluate; an equation is a claim that can turn out true for no values, one value, several values, or every value, as the three equations near the start of this lesson began to show.

Worked example 4 Translate two statements into algebra

Name the unknown number nn in each, then read carefully for a claim that two amounts are equal.

“Twice a number, decreased by five” is a phrase. Twice the number is 2n2n, and decreased by five subtracts 55:

2n−5.2n - 5.

It names a value and asserts nothing, so it is an expression.

“The sum of a number and eight is twenty” is a full statement, and the word is acts as the equals sign. The sum of the number and eight is n+8n + 8, set equal to 2020:

n+8=20.n + 8 = 20.

That equals sign makes it an equation, a sentence you could later solve to find nn.

Common mistakes

Practice

Multiple Choice Questions (MCQ)

Progressively harder sets of questions. Each opens on its own page.

Core practice

Practice problems at the level of the course, to be worked out on paper. Hints one at a time, then the answer or the full worked solution, with your progress kept in this browser.

Core practice Work it out on paper 10 problems Start →
More practice (optional)

Extra sets, as hard as the Challenge set. Each one opens on its own page.

More resources (optional)

Other explanations of this lesson, if you want a second take.

A bit of history (optional)

For centuries the laws in this lesson had no names. Everybody used them, and nobody had a word for them.

The words came late, and they came for a reason. In 1814 a French army officer called Servois chose commutative and distributive. Around 1843 an Irish mathematician, William Rowan Hamilton, supplied associative. Hamilton needed that word in a hurry while building a strange new number system called the quaternions. In it the order of a product suddenly mattered, and there abab and baba can come out different. The commutative law simply fails.

So naming a rule is not tidying up after the real work. Once a rule has a name, you can ask who obeys it. You can also ask who breaks it. Ordinary numbers obey every law in this lesson, with the one exception already noted at the inverse row: zero has no partner for multiplication. That is a fact about ordinary numbers, not a law of thought. The rules you brought from arithmetic turn out to be choices.

One quieter habit reached you from Rene Descartes, writing two centuries earlier. He used letters from the front of the alphabet, aa, bb and cc, for amounts already known, and letters from the far end, xx, yy and zz, for the ones still being hunted. That habit is only a convention, not a rule: this lesson’s own identity 2(x+3)=2x+62(x + 3) = 2x + 6 uses xx for a claim that holds for every number, exactly the job Descartes reserved for aa, bb, and cc. A letter is a hint about how its author was thinking, not proof of which job a variable is doing. What settles that is always the sentence the letter appears in.