← mapBedrock

Exponents & Roots

⚗ Dr. Möbius, from the lab

Multiplication was repeated addition. Now we go up one more floor: exponentiation is repeated multiplication. Easy enough. But then I'm going to ask you something that sounds completely insane — "what the hell is a0a^0? What is a2a^{-2}? What is a1/2a^{1/2}?" — and instead of handing you a memorized answer like some Federation drone, you'll watch those values get forced into existence by one single principle: the laws must survive. Nothing here is arbitrary. Everything is inevitable. Buckle up.

THE BIG IDEA

Powers are repeated multiplication, the exponent laws are proved by counting factors, and the meanings of $a^0$, $a^{-n}$, and $a^{1/n}$ are FORCED — they're the only values that keep the laws true — with roots being the inverse question to powers.

Powers: one more floor up the tower

We've been building a tower. Counting gave us +1+1 repeated. Addition repeated is multiplication: 3×4=4+4+43\times4 = 4+4+4. Now multiplication repeated is exponentiation:

an=aaan factors.a^n = \underbrace{a \cdot a \cdots a}_{n \text{ factors}}.

So 25=22222=322^5 = 2\cdot2\cdot2\cdot2\cdot2 = 32. We call aa the base and nn the exponent, and nn simply counts how many copies of aa get multiplied. For now nn is a positive integer; by the end of this lesson it'll be any rational, and you'll see why we had no choice.

The exponent laws, proved by counting factors

Here's the part the Federation of Boring Textbook Authors hands you as three magic spells to memorize, no explanation included. They're not spells — they're bookkeeping. Watch how this actually works.

Product law: aman=am+na^m \cdot a^n = a^{m+n}. Why? Count the factors. ama^m is mm copies of aa, ana^n is nn copies; mash them together and you have m+nm+n copies:

(aa)m(aa)n=aam+n=am+n.\underbrace{(a\cdots a)}_{m}\cdot\underbrace{(a\cdots a)}_{n} = \underbrace{a\cdots a}_{m+n} = a^{m+n}.

That's the entire proof. 2322=(222)(22)=252^3 \cdot 2^2 = (2\cdot2\cdot2)(2\cdot2) = 2^5.

Power-of-a-power law: (am)n=amn(a^m)^n = a^{mn}. The outer exponent says "nn copies of ama^m," and each copy contributes mm factors, so total factors =mn= m\cdot n:

(am)n=amamn=amn.(a^m)^n = \underbrace{a^m \cdots a^m}_{n} = a^{mn}.

(23)2=2323=26=64(2^3)^2 = 2^3\cdot2^3 = 2^6 = 64. Check: 64=2664 = 2^6. Yes.

Quotient law: aman=amn\dfrac{a^m}{a^n} = a^{m-n} (for a0a\ne 0). Dividing cancels matched factors: a5a2=aaaaaaa=a3\frac{a^5}{a^2} = \frac{a\cdot a\cdot a\cdot a\cdot a}{a\cdot a} = a^3. Three survive, 52=35-2=3.

Counting factors. That's the whole damn thing. Every exponent law, every time.

a0=1a^0 = 1 and an=1/ana^{-n} = 1/a^n are FORCED

Now the part I genuinely love — I have actually shouted about this at beakers. "What is a0a^0?" You can't multiply aa by itself zero times and count — the picture breaks. So we don't guess. We demand the laws keep working and see what a0a^0 is forced to be.

The product law must hold even when one exponent is 00:

ana0=an+0=an.a^n \cdot a^0 = a^{n+0} = a^n.

So a0a^0 is a number that, multiplied by ana^n, gives back ana^n — that's the multiplicative identity from last lesson. Therefore (for a0a\ne 0):

a0=1.a^0 = 1.

Not by decree. By force. Any other value would shatter the product law, and I will not stand for shattered product laws in my lab.

Same trick for negative exponents. What's ana^{-n}? Demand the product law:

anan=an+(n)=a0=1.a^n \cdot a^{-n} = a^{n + (-n)} = a^0 = 1.

So ana^{-n} multiplies ana^n to give 11 — it's the multiplicative inverse of ana^n (straight from Fractions & the Rationals). Therefore:

an=1an.a^{-n} = \frac{1}{a^n}.

So 23=123=182^{-3} = \frac{1}{2^3} = \frac18. A negative exponent isn't "subtraction" — it's "reciprocal." The laws forced it, and you're welcome, because that understanding will save your ass in a dozen future problems.

Fractional exponents: a1/2a^{1/2} must be a\sqrt{a}

Push the same lever. What could a1/2a^{1/2} mean? Demand the power-of-a-power law:

(a1/2)2=a(1/2)2=a1=a.\big(a^{1/2}\big)^2 = a^{(1/2)\cdot 2} = a^1 = a.

So a1/2a^{1/2} is a number that, squared, gives aa. That's exactly what "square root of aa" means. Therefore:

a1/2=a.a^{1/2} = \sqrt{a}.

And a1/n=ana^{1/n} = \sqrt[n]{a} for the same reason — it's the number whose nnth power is aa. Fractional exponents and roots are the same idea. The notation a1/2a^{1/2} is arguably the better one, because it makes the laws automatic: a1/2a1/2=a1/2+1/2=a1=aa^{1/2}\cdot a^{1/2} = a^{1/2+1/2} = a^1 = a. The radical sign is just tradition — pretty tradition, but tradition. The power notation is the real beast.

Roots as the inverse question

Step back. Squaring is an operation: xx2x \mapsto x^2. The square root is its inverse question: x\sqrt{x} asks "what number, squared, gives xx?" — exactly like subtraction asked "what adds back?" and division asked "what multiplies back?". The whole course runs on inverse questions.

But there's a wrinkle. For a positive xx, two numbers square to it: 32=93^2 = 9 and (3)2=9(-3)^2 = 9. So which one is 9\sqrt9? By convention, x\sqrt{x} means the nonnegative one. 9=3\sqrt9 = 3, full stop — not ±3\pm3. (When you want both, you write ±9\pm\sqrt9.) This is a choice, made so that  \sqrt{\ } is an unambiguous answer rather than two answers. Convention, clearly flagged — not a law.

Graph the power and the root side by side; drag the x-probe and watch x\sqrt{x} undo x2x^2 on the right half:

function grapher
x = 1.55x^2 → 2.403sqrt(x) → 1.245

Simplifying radicals, and the landmine

Roots play nicely with products, for nonnegative a,ba, b:

ab=ab.\sqrt{ab} = \sqrt{a}\cdot\sqrt{b}.

(Both square to abab, and both are nonnegative, so they're equal.) This lets you simplify: 12=43=43=23\sqrt{12} = \sqrt{4\cdot3} = \sqrt4\cdot\sqrt3 = 2\sqrt3. Pull out perfect-square factors.

Now the landmine, and I am completely serious: tattoo this on the inside of your eyelids:

a+ba+b.\sqrt{a+b} \ne \sqrt{a} + \sqrt{b}.

Roots do not distribute over addition. Test it: 9+16=25=5\sqrt{9+16} = \sqrt{25} = 5, but 9+16=3+4=7\sqrt9 + \sqrt{16} = 3+4 = 7. Five is not seven. The square root is not linear, and assuming it is will detonate your algebra for years. Products split; sums do not. This error shows up constantly and it's the kind of bullshit that loses points on every exam in this stratum. Remember the difference. Go run the gauntlet.

🔬 SPECIMENS (worked examples)

Worked example 1 — exponent laws, by counting

Simplify x7x2x4\dfrac{x^7 \cdot x^2}{x^4} to a single power of xx (assume x0x \ne 0).

Top first, product law (add exponents — you're merging factor-piles):

x7x2=x7+2=x9.x^7 \cdot x^2 = x^{7+2} = x^9.

Now divide, quotient law (subtract — cancel matched factors):

x9x4=x94=x5.\frac{x^9}{x^4} = x^{9-4} = x^5.

x7x2x4=x5.\frac{x^7 \cdot x^2}{x^4} = x^5.

Sanity check by counting: nine xx's on top, four cancel against the bottom, five survive. The laws are just factor-bookkeeping.

Worked example 2 — negative and zero exponents

Evaluate 50+225^0 + 2^{-2} as a single fraction.

Each piece is forced by the laws, not guessed.

50=15^0 = 1: forced because 5n50=5n5^n\cdot 5^0 = 5^n requires 505^0 to be the multiplicative identity.

22=122=142^{-2} = \dfrac{1}{2^2} = \dfrac14: forced because 2222=20=12^2\cdot 2^{-2} = 2^0 = 1 makes 222^{-2} the reciprocal of 222^2.

Add them with a common denominator (from Fractions & the Rationals):

1+14=44+14=54.1 + \frac14 = \frac44 + \frac14 = \frac54.

50+22=54.5^0 + 2^{-2} = \frac{5}{4}.

Worked example 3 — the trap that's eaten a thousand students

Simplify 50\sqrt{50}, then evaluate 50+14\sqrt{50 + 14}. Watch the difference.

Simplify 50\sqrt{50} by pulling out a perfect-square factor (products split):

50=252=252=52.\sqrt{50} = \sqrt{25 \cdot 2} = \sqrt{25}\cdot\sqrt{2} = 5\sqrt2.

That's legal: 50=25250 = 25\cdot 2 is a product.

Now 50+14\sqrt{50 + 14}. The trap screams "split it into 50+14\sqrt{50}+\sqrt{14}." Do NOT. That's a sum inside the root, and roots don't distribute over sums. Add first:

50+14=64=8.\sqrt{50 + 14} = \sqrt{64} = 8.

Compare with the illegal move: 50+147.07+3.74=10.818\sqrt{50}+\sqrt{14} \approx 7.07 + 3.74 = 10.81 \ne 8. The lesson, burned in: add inside the radical before rooting; only factors may leave the house.

☠ KNOWN HAZARDS

  • The radical landmine: a+b=a+b\sqrt{a+b}=\sqrt{a}+\sqrt{b}. Absolutely false, and this error will haunt you across multiple stratum. 9+16=5\sqrt{9+16}=5 but 3+4=73+4=7. Roots split over products, never over sums. This is the single most expensive mistake in the stratum.

  • Thinking a negative exponent makes a negative number. 23=182^{-3}=\frac18, a positive number. Negative exponent means reciprocal, not "minus." an=1/ana^{-n}=1/a^n. Confusing these two will ruin perfectly good computations.

  • Writing 9=±3\sqrt{9}=\pm3. By convention 9=3\sqrt{9}=3 (the nonnegative root). The equation x2=9x^2=9 has two solutions ±3\pm3, but the symbol 9\sqrt{9} is the one nonnegative value. Keep the symbol and the equation separate.

  • Adding exponents when you should multiply them (or vice versa). aman=am+na^m\cdot a^n=a^{m+n} (multiplying powers \Rightarrow add exponents), but (am)n=amn(a^m)^n=a^{mn} (power of a power \Rightarrow multiply). Count the damn factors and you'll never confuse them.

TL;DR

  • ana^n is nn copies of aa multiplied; the laws aman=am+na^m a^n=a^{m+n}, (am)n=amn(a^m)^n=a^{mn}, am/an=amna^m/a^n=a^{m-n} are proved by counting factors.

  • a0=1a^0=1 and an=1/ana^{-n}=1/a^n are FORCED: they're the only values that keep the product law alive (a0a^0 is the identity, ana^{-n} is the reciprocal of ana^n).

  • a1/n=ana^{1/n}=\sqrt[n]{a} is forced too — (a1/2)2=a(a^{1/2})^2=a means a1/2a^{1/2} squares to aa, i.e. it's a\sqrt{a}.

  • x\sqrt{x} is the inverse question to squaring; positive xx has two square roots, but x\sqrt{x} means the nonnegative one by convention.

  • ab=ab\sqrt{ab}=\sqrt{a}\sqrt{b} (products split), but a+ba+b\sqrt{a+b}\ne\sqrt{a}+\sqrt{b} (sums do NOT) — the landmine.

unlocks