Basis as a Coordinate System (Not Just a Set of Vectors)

When we first learn vectors, coordinates feel absolute.

A vector is

and that’s that.

But linear algebra quietly hides a deeper truth:

Coordinates are not properties of vectors — they are properties of vectors relative to a basis.

This post is about making that idea feel inevitable, not memorized.

The familiar picture: the standard basis

In , we usually work with the standard basis

Geometrically, these vectors define the horizontal and vertical directions.
If we place a vector with tail at the origin, we can write it as

That statement already contains structure:

• the numbers and are coordinates
• the vectors define the grid we are measuring against

This is the coordinate system we implicitly use when we write

So far, nothing surprising.

A different grid, same vector

Now comes the important shift.

Suppose instead we choose two different vectors:

These vectors are not perpendicular, not unit length, and not aligned with the axes —
but they still span the plane.

With respect to this new pair, the same geometric vector can be written as

Suddenly, the coordinates of are no longer .
They are now .

Nothing about the vector changed.
Only the coordinate system did.

Visual intuition: changing the grid, not the vector

Imagine freezing the vector in place.

It stays exactly where it is in the plane. We are not allowed to move it.

Now do the following mental experiment:

1. Draw the standard grid formed by and .

   • horizontal lines every unit
   • vertical lines every unit
   • coordinates make sense because the grid is rectangular

2. Erase the grid, but keep the vector.

3. Draw a new grid using and .

   • lines parallel to
   • lines parallel to
   • the grid is skewed, not orthogonal

Suddenly, the same point sits at the intersection corresponding to .

Nothing moved.
Only the ruler changed.

That’s what a change of basis really is:

• not rotating the vector
• not stretching space
• but replacing the measuring grid underneath it

What actually changed?

This is the key insight:

• A vector is a geometric object.
• A basis is a measuring device.
• Coordinates depend on the measuring device.

When we choose a basis, we are deciding:

• which directions count as “one unit”
• how movement in space is decomposed
• how numbers are assigned meaning

In this sense, a basis is not just a set of vectors —
it is a coordinate system.

Why not any collection of vectors?

Not every set of vectors can play this role.

• A single vector can’t describe the whole plane — some directions are unreachable.
• Too many vectors can describe the same point in multiple ways, destroying uniqueness.

For coordinates to make sense, we need:

1. Existence — every vector must be reachable
2. Uniqueness — every vector must have exactly one coordinate description

These requirements are precisely why bases are defined using:

• spanning
• linear independence

The definition is not arbitrary — it is forced by the idea of coordinates itself.

The takeaway

When you write

you are secretly saying:

“The coordinate vector of relative to the standard basis is .”

Linear algebra begins when we stop treating that statement as obvious
and start asking what happens when the basis changes.

That’s where structure lives.

What’s next

In the next post, we’ll answer a natural question:

Why do we need both spanning and linear independence — and what breaks when one is missing?

That question turns out to explain half of linear algebra.