Foundations · Lesson 2

Operations as transformations

Addition, subtraction, multiplication, and division are not rituals. They are transformations - ways of changing a quantity - and some of them can be undone.

Core ~28 minDeep study ~78 minWhat is an operation?Why can some steps be undone?

Before this: What numbers are

How this idea was born

Everyday life is full of verbs applied to quantities. Combine two piles. Take some away. Scale a recipe for more guests. Split a harvest into equal shares. Undo a mistake when you can, and notice when you cannot. Long before school worksheets, people needed those moves to be reliable: same inputs, same results, and a clear sense of which steps reverse.

Arithmetic traditions wrote laws for those verbs, including awkward cases like zero and debt. In medieval Europe, Leonardo of Pisa (Fibonacci) helped popularize Hindu-Arabic place-value calculation in Liber Abaci (1202), so merchants could add, multiply, and convert with a compact digit system rather than only with abacus habits. The lesson's sidebar figure, Srinivasa Ramanujan, did not invent addition. What he models is the ethic of living inside operations until patterns stop feeling accidental: notebooks packed with structure, often without proofs on the page, because the pages were for him. Steal the obsession with how transformations fit together, not a mysticism myth.

For a modern rereader, this chapter reframes operations as transformations: structured ways a quantity can change, some of which can be undone. You will return here whenever algebra asks you to isolate an unknown, whenever a recipe scales, and whenever a model multiplies an input by a weight. Invertibility is the difference between a step you can reverse and a step that loses information forever.

First principles

An operation is a verb

Think of a machine with an input and an output:

\text{input} \xrightarrow{\text{operation}} \text{output}

Addition by

5

sends every number

x

x + 5

. Multiplication by

3

sends

x

3x

. These are transformations of quantity.

Ask three questions of every operation you meet:

What transformation am I applying?
Does it have an inverse?
What stays invariant?

Addition: combining and shifting

If you have a pile of

a

and a pile of

b

of the same kind, their combine-count is

a + b

. On the number line, addition by

k

is a shift.

Properties that fall out of the idea:

Commutative: $a + b = b + a$
Associative: $(a + b) + c = a + (b + c)$
Identity: $a + 0 = a$

Subtraction asks for the undo of an addition. On the integers, every addition by

k

has an inverse: subtract

k

, or add

-k

Add 5, then subtract 5 - the arrow undoes the arrow.

Why does this work? Because addition by

k

is a shift, and shifts can be shifted back.

Multiplication: scaling and structure

For whole numbers,

3 \times 4

can be read as three copies of four. Better: see it as scaling and structure.

3 × 4 as an array: three rows of four - rearrange and the count stays.

Arrays show why

3 \times 4 = 4 \times 3

: rearrange the grid and the count is unchanged. Commutativity stops being a rule to memorize and becomes something you can see.

Division asks: what input, after scaling by $k$ , produces this output? If

k \neq 0

, multiplication by

k

is invertible:

\text{If } y = kx \text{ and } k \neq 0, \text{ then } x = \frac{y}{k}.

Division by zero has no inverse story: multiplication by

0

crushes every input to

0

. Information about the original

x

is gone.

Distributivity: the rectangle law

The law

a(b+c)=ab+ac

is a rectangle split in two. Factoring and expanding are the same picture in two outfits. When a coupon applies to a subtotal of several items, distributivity is money, not mysticism.

Worked intuition

Example 1 - undo a chain.
Start with

x

, add

3

, multiply by

2

, get

14

. Working backward: divide by

2

to get

7

, subtract

3

to get

x = 4

. Solving equations is invertibility with manners: do the same to both sides, using operations that can be undone.

Example 2 - scale a recipe.
A recipe for

4

becomes a recipe for

10

by multiplying every ingredient by

\tfrac{10}{4} = 2.5

. Halving undoes doubling. Scaling by zero would erase the recipe; that is why nonzero matters.

Example 3 - order that changes the bill.
Discount

10\%

then add

8\%

tax, versus tax then discount: the cash-register order can change what you pay. Commutativity is a claim about order, not a vibe. Sometimes everyday processes refuse to commute.

Example 4 - invariants.
Adding the same amount to both sides of a balanced scale preserves equality. Multiplying both sides by the same nonzero factor preserves equality. Multiplying both sides by zero "preserves" a trivial equality

0=0

while destroying the information you cared about.

Common confusions

"Operations are just worksheets." Procedures are muscle memory. The concept is transformation plus structure (laws) plus invertibility.
"Multiplication always means repeated addition." That reading works for whole numbers. Scaling, rates, and later matrices need a broader verb: multiply transforms by a factor or a linear map.
"Division by zero is a taboo for no reason." The reason is structural: there is no inverse for the crush-to-zero map.
"If it works left-to-right, order never matters." Associativity and commutativity are separate claims. Subtraction and division are not commutative in general: $5-3 \neq 3-5$ .
"Times tables are the concept." Useful memory, thin concept. Ask what stays the same when you rearrange an array.

History and stories

Most school drills treat operations as chores. Flip the question: which everyday processes commute, and which punish you for swapping steps? Baking often fails if you swap "mix wet" and "bake." Putting a file in the trash then emptying the trash does not undo like subtraction undoes addition.

Real world

Double a recipe: multiply by

2

. Halve it: multiply by

\tfrac{1}{2}

. Convert currency: multiply by a rate. Resize an image: scale pixel dimensions by a factor. Each is the same verb - scale - wearing different clothes.

Unit mistakes are operation mistakes: multiply by

1000

when you meant divide, or shift a decimal point, and the transformation lies. Strogatz's phone-bill anecdote (popular exposition, not a primary source) exists to make that visceral.

Open when you want the machine-learning connection. Skip freely.

Field notes, extra examples, and glossary live here when you want more than the core path.

Carry this forward

Operations transform quantities.
Subtraction and division (nonzero) are undoing stories.
Multiplication is scaling and structure.
Next: fractions and ratios - the language of relative size.

Go deeper when curiosity hits - videos, essays, and primary trails.

referenceMacTutor - Srinivasa RamanujanLife, notebooks, and the Hardy correspondence.
video3Blue1Brown - Essence of linear algebra (preview of operations as maps)Later track, but already shows operations as geometric transformations.
videoNumberphile - 1729 and HardyA small window into Ramanujan’s feeling for numerical structure.
articleBerndt - Living with Ramanujan for 40 yearsHow modern mathematicians inhabit his notebooks.
bookSteven Strogatz - The Joy of XRock-pile intuition for why addition and multiplication commute.
articleAlgebrica - Properties of real numbersCommutative, associative, distributive laws with clean structure.

Lesson threads

Real world

Scaling a recipe, converting currency, or resizing an image are multiplications: one quantity transforms another by a factor. Undoing the scale is division - when the factor is not zero.

Mathematician

Srinivasa Ramanujan

1887-1920

Ramanujan did not wait for permission to be intimate with numbers. He filled notebooks the way some people fill diaries - thousands of results, often without proofs on the page, because the notebooks were for him. Carr’s Synopsis taught him a style of compressed discovery; Hardy later gave him a stage. The ethic worth stealing is not mysticism. It is obsession with structure until patterns stop feeling accidental - work that continues when nobody is grading you.

Later bridge · AI / ML

Optional trail - open when useful.

A neural network’s most basic verb is multiply-and-add: weights scale inputs and combine them. Training later asks how to reverse the effect of those operations on a loss - the same invertibility instinct, raised to many layers.

Operations as transformations

How this idea was born

First principles

An operation is a verb

Addition: combining and shifting

Multiplication: scaling and structure

Distributivity: the rectangle law

Worked intuition

Common confusions

History and stories

Real world

Carry this forward

Read more

Check your understanding