When you first meet algebra, a reasonable question is: why the letters? Arithmetic works. You have been adding, subtracting, multiplying, and dividing numbers for years, and they gave you definite answers every time. Now someone writes 3x + 5 on the board and calls that an "expression." What was wrong with plain numbers?

The short answer is that numbers can only describe one case at a time, but the world is full of patterns that apply to every case. Letters let you name "every case" in a single symbol, and that turns out to be one of the most powerful ideas in all of mathematics.

A letter is a slot, not a mystery

The first thing to unlearn is the idea that the letter x is a specific hidden number that you are trying to find. Sometimes it is — when you write 2x + 3 = 11, the x does stand for one particular value (here, 4) that makes the equation true. But that is only one use of letters.

The more fundamental use is this: a letter is a slot that can hold any number from a stated domain. When you write

(x + 1)^2 = x^2 + 2x + 1

and say "this is true for every integer x," you are not hunting for a specific value. You are making a claim about every integer at once. The letter x is a placeholder — drop any integer into the slot, and both sides will give the same answer.

Think of x as an empty box you can fill with any number you like, on the condition that both sides of the equation must be filled with the same number. If you put 7 on the left, you must put 7 on the right. Whatever you choose, the two sides match.

What numbers alone cannot say

Suppose you noticed something true. For every whole number you tested, the square of the next number was exactly the current number squared, plus twice the current number, plus one. You checked:

You now want to tell someone what you found. Without letters, you can only list cases — and there are infinitely many whole numbers, so no list will ever cover them all.

With a letter, you state it once:

(x + 1)^2 = x^2 + 2x + 1 \quad \text{for every whole number } x.

That is the entire claim. One equation, one letter, every case covered. Once you prove it — using the distributive law — it stays proved for every number in the domain forever.

This compression from "infinitely many numeric examples" to "one statement with a letter" is what algebra buys you. It is the same leap as going from listing every multiple of 2 (2, 4, 6, 8, …) to saying "a number is even if it has the form 2k for some integer k." The letter k names all of them at once.

Infinite list of numeric examples collapses into one statement with a letterOn the left, a vertical list shows numeric examples: 1 squared plus 2 plus 1 equals 2 squared, 2 squared plus 4 plus 1 equals 3 squared, 3 squared plus 6 plus 1 equals 4 squared, and so on with dots continuing downward forever. A large arrow labelled collapses to points from the list to a single boxed equation on the right reading x squared plus 2x plus 1 equals (x plus 1) squared, for every integer x. A caption at the bottom says one letter replaces an infinite list. Numbers alone: an endless list 1² + 2(1) + 1 = 2² 2² + 2(2) + 1 = 3² 3² + 2(3) + 1 = 4² 4² + 2(4) + 1 = 5² 5² + 2(5) + 1 = 6² ⋮ (forever) collapses to x² + 2x + 1 = (x + 1)² for every integer x one letter replaces an infinite list
The move from arithmetic to algebra is a move from listing cases to naming the pattern. The letter $x$ does the naming.

A letter can vary on purpose

There is a second job letters do that numbers cannot: represent quantities that genuinely vary.

The perimeter of a square is four times its side length. If you call the side s, the perimeter is 4s. You are not trying to discover some hidden value of s — you are describing a relationship that holds no matter what side length you pick. Build a square with side 3 cm and its perimeter is 12 cm. Build one with side 7 cm and its perimeter is 28 cm. The formula P = 4s captures both — and every other square anyone will ever build.

This is the difference between an unknown (a specific number you are trying to find) and a variable (a quantity that can take many values, and the formula describes the pattern across all of them). The same letter can play either role depending on context. In the equation 4s = 12, the s is an unknown with the single value 3. In the formula P = 4s, the same s is a variable ranging over every possible side length.

Numbers cannot do this job at all. The number 3 is just 3. It cannot "be" the side of every possible square.

Why the ancients invented them

Letters in algebra did not appear from nothing. For most of history, mathematicians wrote everything in words. The Persian mathematician al-Khwarizmi (around 825 CE), whose name gives us the word algorithm, wrote his famous book Al-Jabr entirely in prose. To solve what we would write as x^2 + 10x = 39, he wrote "a square and ten roots equal thirty-nine" — every variable named in words. Leonardo Fibonacci (around 1200 CE) brought this Arabic mathematics to Europe in his Liber Abaci, still writing largely in words and numerical examples. The method worked, but general patterns were hidden inside specific cases.

The big shift came with François Viète in France around 1591. In his In Artem Analyticem Isagoge, Viète used letters — vowels for unknowns, consonants for known quantities — to write general statements about all numbers at once. Suddenly a "sum of two squares" was not the particular number 9 + 16 = 25 but the structure A^2 + B^2. René Descartes refined the notation a few decades later — letters near the end of the alphabet (x, y, z) for unknowns, letters near the start (a, b, c) for constants. After them, you derived the pattern once — say, the quadratic formula x = \frac{-b \pm \sqrt{b^2 - 4ac}}{2a} — and every quadratic for the rest of time was solvable by plugging in numbers for a, b, c. The letters made mathematics general.

A tiny worked example — what letters actually enable

Suppose a shopkeeper in Mumbai sells mangoes for 40 rupees each. You want to know the cost of buying different numbers of mangoes. Without letters, you would compute each case:

With a letter — say n for the number of mangoes — the cost is 40n rupees. One expression, valid for every choice of n. Want to know the cost for 23 mangoes? Put n = 23, get 920 rupees. Want to know how many mangoes you can afford with 500 rupees? Solve 40n = 500 to get n = 12.5, so at most 12.

Now push further. The shopkeeper raises the price to p rupees each. The cost of n mangoes is now np rupees. Two letters, one expression — and it works for any price and any quantity, forever. The expression np is not a number. It is a function, a relationship, a pattern. That is what letters buy you.

Common confusions

Where this takes you next

Once you accept that a letter is a slot that ranges over a domain, the rest of algebra becomes natural. An expression like 3x + 5 is a recipe — put any number in the slot, get a number out. An equation like 3x + 5 = 14 is a question — which values in the slot make both sides match? An identity like (x + 1)^2 = x^2 + 2x + 1 is a promise — every value in the slot makes both sides match. All three ideas depend entirely on the letter meaning "any value, not a particular one."

Without letters, none of this is sayable. You can do arithmetic forever and never state a single general law. With letters, one line of algebra can say more than every arithmetic textbook in the world put together.

Head back to Algebraic Expressions to see how the vocabulary — terms, coefficients, degree — is built on top of this one idea: that a letter is a slot waiting to be filled.