1. Coordinates as Scalars
Definition 1.1 – Standard basis vectors
In the xy-plane there are two special unit vectors:
Key idea – Reading coordinates as scaling instructions
Each coordinate is a scalar that scales a basis vector; the vector is the sum of the scaled basis vectors.
Worked example.
- stretches to three times its length.
- flips and stretches it by 2.
- Adding the two scaled vectors gives the vector.
Definition 1.2 – Basis (informal)
and together form the basis of the standard coordinate system: they are the things that the coordinates actually scale.
2. Choosing a Different Basis
Nothing forces us to use and . Pick two other vectors — e.g. one pointing up-right and one pointing down-right — and use pairs of scalars to scale them instead.
- For most such pairs, every 2D vector can still be reached.
- This gives a valid but different translation between number pairs and arrows.
⚠️ Pitfall: Whenever you write a vector as numbers, you have implicitly chosen a basis. The same arrow has different coordinates in different bases (full treatment in Chapter 13).
3. Linear Combinations
Definition 3.1 – Linear combination
Scaling vectors and adding the results:
is a linear combination of and .
Why “linear”?
Not the etymology, but a useful picture: fix one scalar and let the other vary freely — the tip of traces a straight line.
4. Span
Definition 4.1 – Span
The span of a set of vectors is the set of all their linear combinations — everything reachable using only vector addition and scalar multiplication.
Span of two vectors in 2D
| Situation | Span |
|---|---|
| point in different directions (most pairs) | All of the 2D plane |
| line up (one is a multiple of the other) | A single line through the origin |
| Just the origin |
Vectors as points
- A single vector → think of it as an arrow.
- A collection of vectors → think of each as the point at its tip (tail still at the origin).
So “all vectors on a line” = the line itself, and “all 2D vectors” = the whole plane.
Span in 3D
| Vectors | Typical span |
|---|---|
| Two non-parallel 3D vectors | A flat plane through the origin |
| Add a third vector on that plane | Still the same plane (no new reach) |
| Add a third vector off that plane | All of 3D space |
Picture: as you scale the third vector, it slides the plane spanned by the first two through all of space — three free scalars give access to three dimensions.
Worked example. , span the xy-plane. Adding changes nothing; adding instead unlocks all of .
5. Linear Dependence and Independence
Definition 5.1 – Linearly dependent
A set of vectors is linearly dependent if at least one can be removed without shrinking the span — equivalently, one of them is a linear combination of the others:
Definition 5.2 – Linearly independent
Each vector adds a new dimension to the span; none is a combination of the others:
| Dependent | Independent | |
|---|---|---|
| Redundant vector? | Yes | No |
| Two 2D vectors | Collinear | Not collinear |
| Three 3D vectors | All lie in one plane (or line) | Span all of 3D |
⚠️ Pitfall: Any set containing is dependent — the zero vector never adds anything to the span.
6. Basis (Technical Definition)
Definition 6.1 – Basis of a space
A basis of a space is a set of linearly independent vectors that span that space.
Why it makes sense:
- Spans the space → every vector can be written as a combination (coordinates exist).
- Independent → no redundancy, so those coordinates are unique.
Exercises: Explain in your own words why “linearly independent + spans the space” is the right definition of a basis; then decide whether form a basis of (practice set on the lesson page).