This deck covers the eight vector-space axioms over a field, the consequences those axioms immediately force, and the zoo of genuine examples: tuples, matrices, polynomials, functions, the solution sets of homogeneous systems, and C regarded over different fields. It targets the misconceptions that a vector is an arrow in R^n, that any set of vectors is a space, that closure under scalar multiplication comes for free, and that the choice of scalar field does not affect dimension.
Subject: Foundations of Higher Mathematics · 109 slides · symbolic lesson
Open the interactive version of this deck · Homework for this lesson
Objectives
By the end of this deck you can:
1. State the eight vector-space axioms over a field and say which family each belongs to.
2. Derive the basic consequences (uniqueness of the zero vector, the zero-scalar rule, the negation rule) directly from the axioms.
3. Recognize the standard examples: tuples, matrices, polynomials, functions, sequences, and solution sets of homogeneous systems.
4. Test a candidate set and decide whether it is a vector space, catching the failures of closure and the missing zero vector.
5. Explain how the chosen field of scalars changes the space, using the complex numbers over the reals versus over themselves.
Warm-up
Discussion prompt
Before we open Vector Spaces: Axioms & Examples: without looking back, what was the main idea of Fields, Characteristic & Finite Fields, and what could you do by the end of it that you could not do before?
Hint: One sentence for the idea, one for the skill. If the second one is blank, that is the part to revisit.
Answer:
Builds the field concept as the top of the ring ladder: division everywhere, no zero divisors, and the characteristic that is forced to be zero or prime. Constructs the smallest genuinely new finite field F_4 from scratch with full addition and multiplication tables, and hammers the central trap that the field with four elements is NOT the ring of integers modulo four.
Concept
A vector space is not one object. It is a package: a set, a field of scalars, and two operations that fit together by rule.
Write the set of vectors and the field of scalars, with an addition on vectors and a scaling by field elements.
\[ (V, +, \cdot) \text{ over a field } \mathbb{F} \]
vector space — A set V with an addition on V and a scalar multiplication by elements of a field F, satisfying the eight axioms. The elements of V are called vectors; the elements of F are called scalars.
Counterexample
Discussion prompt
A vector space is not one object. It is a package: a set, a field of scalars, and two operations that fit together by rule.
That is stated as though it always holds. Do one of two things: produce a case where it fails, or say precisely what rules such a case out. "It just does" is not on the menu.
Hint: Hunt at the extremes first — zero, one, negative, empty, equal. If every extreme survives, the reason they survive is the proof.
Answer:
Write the set of vectors and the field of scalars, with an addition on vectors and a scaling by field elements.
Intuition
Forget arrows for a moment. The word vector names a role, not a shape.
If you have a collection of things where adding two of them and stretching one by a number both make sense and behave sanely, those things are vectors and the collection is a candidate vector space.
Polynomials, matrices, and functions all pass this test. That is why the same theorems govern all of them at once.
Analogy
Discussion prompt
Explain A vector is anything you can add and scale by analogy to something with no Foundations of Higher Mathematics in it at all — a queue, a recipe, a map, a bank balance, whatever fits. Then say where your analogy breaks.
Hint: An analogy that never breaks is not an analogy, it is the same idea wearing a hat. Find the seam — that is the part that is actually new.
Answer:
Forget arrows for a moment. The word vector names a role, not a shape.
Concept
The scalars are not arbitrary. They form a field: you can add, subtract, multiply, and divide by anything nonzero.
\[ \mathbb{F} \in \{ \mathbb{Q}, \mathbb{R}, \mathbb{C}, \mathbb{F}_p, \dots \} \]
Division is what later lets every vector space have a basis. Drop division and you get a module over a ring, which is much wilder. Fields are the well-behaved case.
Explain it
Discussion prompt
Explain The scalars come from a field to a student a year behind you. No notation, no jargon they have not met — and it still has to be true.
Hint: If your explanation needs a symbol they have never seen, you are describing the notation rather than the idea.
Answer:
The scalars are not arbitrary. They form a field: you can add, subtract, multiply, and divide by anything nonzero.
Concept
The package carries exactly two operations, and each must land back inside the set.
Vector addition takes two vectors to a vector.
\[ + : V \times V \to V \]
Scalar multiplication takes a scalar and a vector to a vector.
\[ \cdot : \mathbb{F} \times V \to V \]
The arrows landing in V is closure. It is an axiom, not a freebie, and it is the one most often quietly assumed.
Concept
The first axiom on addition: grouping does not matter.
\[ (u + v) + w = u + (v + w) \quad \text{for all } u, v, w \in V \]
Concept
Order of addition does not matter either.
\[ u + v = v + u \quad \text{for all } u, v \in V \]
Associativity and commutativity together say the vectors form an abelian group under addition, once we add an identity and inverses.
Concept
There is a distinguished vector that changes nothing when added.
\[ \exists\, 0 \in V \ \text{ such that }\ v + 0 = v \ \text{ for all } v \in V \]
zero vector — The additive identity of V. It is a vector, not the scalar zero, though the two are closely linked. We will prove it is unique.
Concept
Each vector can be undone by adding a partner that returns you to zero.
\[ \forall v \in V \ \exists\, (-v) \in V \ \text{ such that }\ v + (-v) = 0 \]
With associativity, commutativity, a zero, and inverses, the pair consisting of V and its addition is a full abelian group.
Intuition
Do not memorize eight unrelated rules. Four of them are one familiar package.
Associativity, commutativity, a zero, and inverses mean addition alone already makes V an abelian group, exactly like the integers under addition.
The remaining four axioms are the new content: they say the scalar multiplication cooperates with that group and with the field.
Concept
Scaling by one number and then another equals scaling by their product.
\[ a\,(b\,v) = (ab)\,v \quad \text{for all } a, b \in \mathbb{F},\ v \in V \]
The product inside the parentheses on the right is multiplication in the field; the products on the left are scalings of a vector. The axiom links the two.
Concept
The multiplicative identity of the field acts as the identity on vectors.
\[ 1 \cdot v = v \quad \text{for all } v \in V \]
This axiom is easy to overlook but essential. Without it, scaling could collapse every vector to zero and still satisfy the other seven rules.
Concept
A single scalar spread across a sum of vectors splits over the sum.
\[ a\,(u + v) = a\,u + a\,v \quad \text{for all } a \in \mathbb{F},\ u, v \in V \]
Concept
A sum of scalars acting on one vector splits over the scalar sum.
\[ (a + b)\,v = a\,v + b\,v \quad \text{for all } a, b \in \mathbb{F},\ v \in V \]
The two distributive laws look symmetric but are different: one sum lives in V, the other in F. Both are required.
Picture it
Figure (svg): Three labeled boxes: an abelian-group box holding four addition axioms, a compatibility box holding scalar associativity and the unit law, and a distributive box holding the two distributive laws.
Discussion prompt
Read the picture before the words. What is this showing, and what is the one thing it is built to make obvious? Commit to an answer, then read on.
Hint: Name the parts, then say what changes between them — and if nothing changes, say what is being held still.
Answer:
Group them so they are easy to reconstruct on an exam.
Intuition
Group them so they are easy to reconstruct on an exam.
Four say addition is an abelian group. One says scalar multiplication is associative with field multiplication. One is the unit law. Two are the distributive laws.
Figure (svg): Three labeled boxes: an abelian-group box holding four addition axioms, a compatibility box holding scalar associativity and the unit law, and a distributive box holding the two distributive laws.
Ranking
Put in order
Put the moves of Worked example: the plane is a vector space into the order they have to happen.
Why: These are the moves of the worked example in the order it makes them, and each one is set up by the one before it. The pair of zeros leaves any pair unchanged, since adding zero in each coordinate does nothing.
Worked example
Take the set of ordered pairs of real numbers with componentwise addition and scaling. Confirm two representative axioms.
\[ V = \mathbb{R}^2, \quad (x_1,x_2)+(y_1,y_2)=(x_1+y_1,\ x_2+y_2) \]
Exhibit the zero vector
Why: The pair of zeros leaves any pair unchanged, since adding zero in each coordinate does nothing.
\[ (x_1,x_2)+(0,0) = (x_1,x_2) \]
Exhibit the negative of a vector
Why: Negating each coordinate returns the zero vector, so inverses exist.
\[ (x_1,x_2)+(-x_1,-x_2) = (0,0) \]
Check one distributive law on a scalar sum
Why: Reduce both sides to coordinates and use the distributive law of the real numbers in each slot.
\[ (a+b)(x_1,x_2) = ((a+b)x_1,\,(a+b)x_2) = (ax_1+bx_1,\ ax_2+bx_2) \]
Verify the two sides agree
Why: The right-hand expression is exactly the coordinate form of a(x1,x2)+b(x1,x2), so the distributive axiom holds; the remaining axioms follow identically from the field laws of R.
\[ a(x_1,x_2)+b(x_1,x_2) = (ax_1+bx_1,\ ax_2+bx_2) \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the plane is a vector space", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: Reduce both sides to coordinates and use the distributive law of the real numbers in each slot.
Concept
The eight axioms are the entire contract. Anything true in every vector space must be squeezed out of them alone.
Three consequences come up constantly: the zero vector is unique, the scalar zero annihilates every vector, and scaling by negative one gives the additive inverse. We prove all three.
Step zero
Discussion prompt
Worked example: the zero vector is unique — before any calculation: what is the plan? Name the moves in order, in plain English, without doing the arithmetic.
Hint: It starts with: Use the first zero on the second
Answer:
Worked example
The axiom promises at least one zero vector. Suppose two vectors both act as a zero and show they must coincide.
\[ \text{Suppose } 0 \text{ and } 0' \text{ both satisfy } v + 0 = v \text{ and } v + 0' = v. \]
Use the first zero on the second
Why: Because the first is a zero, adding it to any vector leaves it fixed; apply that with the second zero as the vector.
\[ 0' + 0 = 0' \]
Use the second zero on the first
Why: By commutativity the sum can be reordered, and the second zero fixes the first vector.
\[ 0' + 0 = 0 + 0' = 0 \]
Verify the two expressions collapse
Why: Both computations equal the same sum, so the two candidate zeros are equal; the zero vector is unique.
\[ 0' = 0' + 0 = 0 \ \Rightarrow\ 0' = 0 \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the zero vector is unique", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: Both computations equal the same sum, so the two candidate zeros are equal; the zero vector is unique.
Estimation
Predict first
Distinguish the scalar zero from the vector zero, then prove scaling by the scalar zero always lands on the vector zero.
Commit before you compute: what does Worked example: the scalar zero kills every vector come out to? A rough magnitude and the right form is enough — the point is to have something concrete to be wrong about.
Correct: Verify with a concrete vector
Why: A prediction you can defend turns the computation into a check rather than a leap of faith — and an answer that contradicts it is caught on the spot. In R^2 the scalar zero times any pair gives the pair of zeros, matching the general result.
Worked example
Distinguish the scalar zero from the vector zero, then prove scaling by the scalar zero always lands on the vector zero.
\[ \text{Claim: } 0_{\mathbb{F}} \cdot v = 0_V \text{ for every } v \in V. \]
Split the scalar zero as a sum
Why: In the field, zero plus zero is zero, so scaling by it equals scaling by that sum; then apply the distributive law over scalar addition.
\[ 0v = (0+0)v = 0v + 0v \]
Add the negative of that vector to both sides
Why: The vector 0v has an additive inverse; adding it cancels one copy on the right and yields the zero vector on the left.
\[ 0v + (-(0v)) = (0v + 0v) + (-(0v)) \]
Simplify using associativity and inverses
Why: Regroup on the right so one copy cancels, leaving a single 0v; the left side is the zero vector.
\[ 0_V = 0v \]
Verify with a concrete vector
Why: In R^2 the scalar zero times any pair gives the pair of zeros, matching the general result.
\[ 0 \cdot (3, -5) = (0, 0) = 0_V \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the scalar zero kills every vector", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: In R^2 the scalar zero times any pair gives the pair of zeros, matching the general result.
Missing information
Discussion prompt
The negation of a vector is defined by the inverse axiom. Show it is always produced by scaling by the scalar negative one.
What do you need to know — or decide — before the first line can be written? List everything the problem has to hand you.
Hint: Anything you would have to invent to get started is a thing the problem must supply.
Answer:
Rewrite v as one times v using the unit axiom, then combine with the distributive law over scalar addition.
Worked example
The negation of a vector is defined by the inverse axiom. Show it is always produced by scaling by the scalar negative one.
\[ \text{Claim: } (-1)\cdot v = -v \text{ for every } v \in V. \]
Add v to the candidate
Why: Rewrite v as one times v using the unit axiom, then combine with the distributive law over scalar addition.
\[ v + (-1)v = 1\cdot v + (-1)v = (1 + (-1))v \]
Collapse the scalar sum
Why: In the field, one plus negative one is zero, and scaling by the scalar zero gives the zero vector by the previous result.
\[ (1 + (-1))v = 0 \cdot v = 0_V \]
Read off the inverse
Why: The candidate added to v yields the zero vector, so by uniqueness of inverses it is exactly the negative of v.
\[ v + (-1)v = 0_V \ \Rightarrow\ (-1)v = -v \]
Verify in the plane
Why: Scaling a concrete pair by negative one negates each coordinate, matching the additive inverse.
\[ (-1)(3, -5) = (-3, 5) = -(3,-5) \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "negative one gives the inverse", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: Scaling a concrete pair by negative one negates each coordinate, matching the additive inverse.
Concept
Two zeros, two kinds of object, one bar between them. Keep them straight or the proofs blur.
Scalars live in the field and can be divided. Vectors live in the space and generally cannot be multiplied together at all.
| object | lives in | can you divide? |
|---|---|---|
| scalar | the field F | yes, except by zero |
| vector | the space V | no vector division |
Comparison
Comparison matrix
From Scalars and vectors are different citizens: refill the lives in column from what you know. The rest of the table is as it appeared.
| object | lives in | can you divide? |
|---|---|---|
| scalar | the field F | yes, except by zero |
| vector | the space V | no vector division |
Trap
Believing vector always means an arrow in two or three dimensions, so anything without a picture cannot be a vector.
Under this view a polynomial or a function could never be a vector, and linear algebra would not apply to them.
\[ 3x^2 - x + 4 \ \text{``is not a vector''} \]
A vector is any element of a set that satisfies the axioms. Arrows in the plane are one example among many.
Polynomials add and scale by the field, so they are genuine vectors; the axioms, not the picture, decide.
\[ (3x^2 - x + 4) + 2(x^2 + 1) = 5x^2 - x + 6 \in \mathcal{P} \ \checkmark \]
Concept
The first family of examples is fixed-length lists of field elements, added and scaled slot by slot.
\[ \mathbb{F}^n = \{ (x_1, \dots, x_n) : x_i \in \mathbb{F} \} \]
Every axiom reduces to a field law applied in each coordinate, so all of these are vector spaces for free. This is the model every later example imitates.
Concept
Fix the shape. All matrices of one fixed size, added entry by entry and scaled entry by entry, form a space.
\[ M_{m \times n}(\mathbb{F}) = \{ \text{all } m \times n \text{ arrays over } \mathbb{F} \} \]
Here vector addition is matrix addition. Matrix multiplication is irrelevant to the vector-space structure; it is extra data the space does not require.
Step zero
Discussion prompt
Worked example: two-by-two matrices — before any calculation: what is the plan? Name the moves in order, in plain English, without doing the arithmetic.
Hint: It starts with: Exhibit the zero matrix
Answer:
Worked example
Confirm the set of real two-by-two matrices is a vector space by checking the zero and a distributive law.
\[ V = M_{2\times 2}(\mathbb{R}), \quad \text{addition and scaling entrywise} \]
Exhibit the zero matrix
Why: The matrix of all zeros added to any matrix leaves every entry unchanged, so it is the additive identity.
\[ \begin{pmatrix} a & b \\ c & d \end{pmatrix} + \begin{pmatrix} 0 & 0 \\ 0 & 0 \end{pmatrix} = \begin{pmatrix} a & b \\ c & d \end{pmatrix} \]
Check distributivity over a vector sum
Why: Scaling a sum of matrices scales each entry of the entrywise sum, and each entry obeys the distributive law of R.
\[ \lambda\left( A + B \right) \ \text{has } (i,j)\text{-entry } \lambda(a_{ij}+b_{ij}) = \lambda a_{ij} + \lambda b_{ij} \]
Verify against the entrywise sum
Why: The result is exactly the entry of the matrix lambda A plus lambda B, so the axiom holds; every remaining axiom is an entrywise field law.
\[ \lambda(A+B) = \lambda A + \lambda B \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "two-by-two matrices", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: Scaling a sum of matrices scales each entry of the entrywise sum, and each entry obeys the distributive law of R.
Concept
A vector space is never empty: an axiom demands a zero vector. The extreme case is the space that contains only that zero.
\[ V = \{ 0 \}, \quad 0 + 0 = 0, \quad a \cdot 0 = 0 \]
Every axiom holds trivially. This zero space is the additive identity object of the whole theory, the empty-looking case worth naming so it never surprises you.
Concept
Fix a ceiling on the degree. Polynomials with degree no larger than that ceiling, added and scaled as usual, form a space.
\[ \mathcal{P}_n = \{ a_0 + a_1 x + \dots + a_n x^n : a_i \in \mathbb{F} \} \]
Adding two such polynomials keeps the degree at or below the ceiling, and so does scaling, so the set is closed under both operations.
Concept
Drop the ceiling. Every polynomial of any degree, together, is also a vector space.
\[ \mathcal{P} = \bigcup_{n \ge 0} \mathcal{P}_n \]
This one is infinite-dimensional: no finite list of polynomials can produce every polynomial by combination. It is the first natural space that outruns any tuple space.
Intuition
A one-word change breaks everything. Degree at most n and degree exactly n behave completely differently.
The at-most set includes the zero polynomial and is closed under addition. The exactly-n set excludes zero and can lose its top term when you add, so it leaks out of itself.
Fill the middle
Fill in the blanks
From Worked example: degree at most two is a space — finish the line. Write what belongs on the right of the equals sign before you look.
\mathcal\{ a_0 + a_1 x + a_2 x^2 : a_i \in \mathbb{R} \}_2 = ___
Why: Producing the right-hand side unprompted is the difference between recognising this line and being able to use it. The zero polynomial has all coefficients zero, degree conventionally not exceeding two, so it belongs to the set and is the additive identity.
Worked example
Take real polynomials of degree at most two. Check the zero vector and closure under addition.
\[ \mathcal{P}_2 = \{ a_0 + a_1 x + a_2 x^2 : a_i \in \mathbb{R} \} \]
Locate the zero vector
Why: The zero polynomial has all coefficients zero, degree conventionally not exceeding two, so it belongs to the set and is the additive identity.
\[ 0 = 0 + 0x + 0x^2 \in \mathcal{P}_2 \]
Add two general members
Why: Add coefficient by coefficient; each coefficient stays real and no term of degree beyond two appears.
\[ (a_0+a_1x+a_2x^2)+(b_0+b_1x+b_2x^2) = (a_0+b_0)+(a_1+b_1)x+(a_2+b_2)x^2 \]
Verify the sum stays in the set
Why: The result has degree at most two with real coefficients, so it lies in the set; scaling behaves the same way and the remaining axioms are coefficientwise field laws.
\[ (a_0+b_0)+(a_1+b_1)x+(a_2+b_2)x^2 \in \mathcal{P}_2 \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "degree at most two is a space", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: The result has degree at most two with real coefficients, so it lies in the set; scaling behaves the same way and the remaining axioms are coefficientwise field laws.
Estimation
Predict first
Now demand the degree be exactly two. Test whether this set is a vector space.
Commit before you compute: what does Worked example: degree exactly two fails come out to? A rough magnitude and the right form is enough — the point is to have something concrete to be wrong about.
Correct: Verify the failure is real
Why: A prediction you can defend turns the computation into a check rather than a leap of faith — and an answer that contradicts it is caught on the spot. The set lacks a zero vector and is not closed under addition, so at least two axioms fail; it is not a vector space.
Worked example
Now demand the degree be exactly two. Test whether this set is a vector space.
\[ S = \{ a_0 + a_1 x + a_2 x^2 : a_2 \neq 0 \} \]
Search for the zero vector
Why: The zero polynomial has a leading coefficient of zero, so it violates the exactly-two condition and is not in the set.
\[ 0 \notin S \ \text{ since its } x^2 \text{ coefficient is } 0 \]
Also test closure under addition
Why: Two degree-two polynomials can cancel their leading terms, dropping the degree below two and escaping the set.
\[ (x^2 + x) + (-x^2 + 1) = x + 1 \notin S \]
Verify the failure is real
Why: The set lacks a zero vector and is not closed under addition, so at least two axioms fail; it is not a vector space.
\[ S \text{ is not a vector space} \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "degree exactly two fails", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: The set lacks a zero vector and is not closed under addition, so at least two axioms fail; it is not a vector space.
Trap
Checking only that a set is closed under addition, then declaring it a vector space.
Take the integer pairs inside the real plane. They are closed under addition, so the shortcut says vector space.
\[ \mathbb{Z}^2 \subseteq \mathbb{R}^2, \quad (1,0)+(0,1)=(1,1) \in \mathbb{Z}^2 \]
You must also close under scalar multiplication by every field element, not just addition.
Scaling an integer pair by a real scalar leaves the integers immediately, so the set is not a real vector space.
\[ \tfrac{1}{2}\,(1,0) = (\tfrac{1}{2}, 0) \notin \mathbb{Z}^2 \ \checkmark \]
Break the constraint
Discussion prompt
The rule this trap just fixed:
You must also close under scalar multiplication by every field element, not just addition.
Now break it on purpose. Build a case that violates it and follow the consequences until something visibly fails. Where does the failure first show up — and would you have noticed it if you had not been looking?
Hint: The dangerous rules are the ones whose violation still produces an answer. If yours fails loudly, try to find one that fails quietly.
Concept
Fix any set of inputs. All functions from that set into the field, added and scaled pointwise, form a vector space.
\[ \mathcal{F}(X, \mathbb{F}) = \{ f : X \to \mathbb{F} \}, \quad (f+g)(x) = f(x) + g(x) \]
Scaling is defined the same way, one output at a time. Every axiom is checked at each point separately, where it is just a field law.
Intuition
Think of a function as a tuple with one slot for every input, possibly infinitely many slots.
Adding functions adds the values in each slot; scaling scales each slot. This is exactly the tuple recipe, stretched to allow an infinite index set.
Tuples, sequences, and functions are the same idea at three sizes of index set: finite, countable, and arbitrary.
Ranking
Put in order
Put the moves of Worked example: the function space into the order they have to happen.
Why: These are the moves of the worked example in the order it makes them, and each one is set up by the one before it. The function sending every input to zero is the additive identity, since adding it changes no output value.
Worked example
Verify functions from a set into the reals form a vector space by exhibiting the zero and checking a distributive law pointwise.
\[ V = \mathcal{F}(X, \mathbb{R}) \]
Name the zero function
Why: The function sending every input to zero is the additive identity, since adding it changes no output value.
\[ 0(x) = 0 \ \text{ for all } x, \quad (f + 0)(x) = f(x) \]
Check a distributive law at a point
Why: Evaluate both sides at an arbitrary input; the values are real numbers, so the real distributive law applies in that single slot.
\[ \big(a(f+g)\big)(x) = a\big(f(x)+g(x)\big) = a f(x) + a g(x) \]
Verify the functions are equal
Why: The two functions agree at every input, hence are the same function; every other axiom is likewise a pointwise field law.
\[ a(f+g) = af + ag \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the function space", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: Evaluate both sides at an arbitrary input; the values are real numbers, so the real distributive law applies in that single slot.
Concept
A sequence is just a function whose inputs are the natural numbers, so sequence spaces are a special case of function spaces.
\[ \mathbb{F}^{\mathbb{N}} = \{ (a_0, a_1, a_2, \dots) : a_i \in \mathbb{F} \} \]
Add termwise, scale termwise. These infinite-dimensional spaces are the natural home of series, generating functions, and shift operators.
Concept
Consider a linear system whose right-hand side is all zeros. Its full set of solutions is a vector space.
\[ \{ x \in \mathbb{F}^n : A x = 0 \} \]
Homogeneous means the constant column is zero. That single fact is exactly what supplies the zero vector and the closure the axioms demand.
Intuition
Picture the solution set geometrically. A homogeneous system always has the origin as a solution, since the matrix sends zero to zero.
So the solution set is a flat sheet through the origin: a line, a plane, or a higher analog that contains the zero vector and is closed under combinations.
Attach a nonzero constant column and the sheet lifts off the origin. It becomes a parallel copy that no longer contains zero.
Hypothesis
Predict first
Worked example: the homogeneous solution set is about to be worked. State your hypothesis first: which rule or definition decides this one, and what is the first move it forces? Then watch whether the example agrees with you.
Correct: Confirm zero is a solution
Why: The matrix maps the zero vector to the zero vector, so zero satisfies the system and W contains a zero vector.
A hypothesis you wrote down is falsifiable; a vague sense of how it will go is not. If the example opens somewhere else, that gap is the thing worth chasing.
Worked example
Show the solutions of a homogeneous system are closed under addition and scaling, so they form a vector space.
\[ W = \{ x : A x = 0 \} \]
Confirm zero is a solution
Why: The matrix maps the zero vector to the zero vector, so zero satisfies the system and W contains a zero vector.
\[ A \cdot 0 = 0 \ \Rightarrow\ 0 \in W \]
Add two solutions
Why: Apply the matrix to the sum; by linearity of the matrix action it distributes, and each piece is zero.
\[ A(x + y) = Ax + Ay = 0 + 0 = 0 \]
Scale a solution
Why: The matrix action pulls scalars out, so a scaled solution maps to a scaled zero, which is still zero.
\[ A(\lambda x) = \lambda (A x) = \lambda \cdot 0 = 0 \]
Verify closure holds
Why: The set contains zero and is closed under addition and scaling, so it inherits every axiom from the ambient tuple space; it is a vector space.
\[ x, y \in W,\ \lambda \in \mathbb{F} \ \Rightarrow\ x+y \in W,\ \lambda x \in W \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the homogeneous solution set", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: The set contains zero and is closed under addition and scaling, so it inherits every axiom from the ambient tuple space; it is a vector space.
Step zero
Discussion prompt
Worked example: the inhomogeneous set is not a space — before any calculation: what is the plan? Name the moves in order, in plain English, without doing the arithmetic.
Hint: It starts with: Test the zero vector
Answer:
Worked example
Now give the same system a nonzero right-hand side and test the solution set.
\[ U = \{ x : A x = b \}, \quad b \neq 0 \]
Test the zero vector
Why: The matrix sends zero to zero, but the target is a nonzero vector, so zero is not a solution and cannot be in the set.
\[ A \cdot 0 = 0 \neq b \ \Rightarrow\ 0 \notin U \]
Test closure under addition
Why: Adding two solutions doubles the right-hand side, so the sum solves a different equation and leaves the set.
\[ A(x + y) = b + b = 2b \neq b \]
Verify it fails the axioms
Why: The set has no zero vector and is not closed, so it is not a vector space; it is a translate of the homogeneous solution space.
\[ U = x_0 + W \ \text{(an affine set, not a space)} \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the inhomogeneous set is not a space", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: The set has no zero vector and is not closed, so it is not a vector space; it is a translate of the homogeneous solution space.
Trap
Thinking any flat sheet, such as a line that misses the origin, is a vector space because it looks straight.
Consider the line of points whose coordinates sum to one. Adding two of its points breaks the condition.
\[ (1,0)+(0,1) = (1,1), \quad 1 + 1 = 2 \neq 1 \]
A vector space must contain the zero vector and be closed. Sheets that miss the origin fail both and are called affine, not linear.
The line through the origin whose coordinates sum to zero is closed and contains zero, so that one is a genuine subspace.
\[ (1,-1)+(2,-2) = (3,-3), \quad 3 + (-3) = 0 \ \checkmark \]
Concept
A vector space is a set together with a chosen field. Change the field and you have changed the space, even if the underlying set of vectors is identical.
The complex numbers are the sharp example. The same points can be regarded as a space over the reals or over the complex numbers, and the two are genuinely different.
\[ \mathbb{C} \text{ over } \mathbb{R} \quad \text{vs} \quad \mathbb{C} \text{ over } \mathbb{C} \]
Picture it
Figure (svg): The complex plane with the real axis and the imaginary axis drawn, the point 3 plus 2i marked, and dashed lines showing its real part 3 and imaginary part 2 as two real coordinates.
Discussion prompt
Read the picture before the words. What is this showing, and what is the one thing it is built to make obvious? Commit to an answer, then read on.
Hint: Name the parts, then say what changes between them — and if nothing changes, say what is being held still.
Answer:
Coordinates count how many independent directions you need to reach every vector using the scalars you are allowed.
Intuition
Coordinates count how many independent directions you need to reach every vector using the scalars you are allowed.
With only real scalars, reaching every complex number needs two directions, one real and one imaginary. With complex scalars, a single direction suffices, since any complex number is a complex multiple of one.
Figure (svg): The complex plane with the real axis and the imaginary axis drawn, the point 3 plus 2i marked, and dashed lines showing its real part 3 and imaginary part 2 as two real coordinates.
Missing information
Discussion prompt
Regard the complex numbers with real scalars only. Find how many coordinates each vector needs.
What do you need to know — or decide — before the first line can be written? List everything the problem has to hand you.
Hint: Anything you would have to invent to get started is a thing the problem must supply.
Answer:
Every complex number splits into a real part and an imaginary part, each a real multiple of one and of the imaginary unit.
Worked example
Regard the complex numbers with real scalars only. Find how many coordinates each vector needs.
\[ V = \mathbb{C}, \quad \text{scalars from } \mathbb{R} \]
Write a general vector in a real combination
Why: Every complex number splits into a real part and an imaginary part, each a real multiple of one and of the imaginary unit.
\[ a + bi = a\cdot 1 + b\cdot i, \quad a, b \in \mathbb{R} \]
Argue the two directions are independent over R
Why: No real multiple of one equals the imaginary unit, so neither of the two building blocks is a real scalar multiple of the other.
\[ s\cdot 1 = i \ \text{ has no solution } s \in \mathbb{R} \]
Verify the count
Why: Every vector is a unique real combination of two independent vectors, so the real dimension is two; the coordinates of a plus b i are the pair a, b.
\[ \dim_{\mathbb{R}} \mathbb{C} = 2 \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "C as a real vector space", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: Every vector is a unique real combination of two independent vectors, so the real dimension is two; the coordinates of a plus b i are the pair a, b.
Estimation
Predict first
Now allow complex scalars on the same set and recount the coordinates.
Commit before you compute: what does Worked example: C as a complex vector space come out to? A rough magnitude and the right form is enough — the point is to have something concrete to be wrong about.
Correct: Verify the count
Why: A prediction you can defend turns the computation into a check rather than a leap of faith — and an answer that contradicts it is caught on the spot. One vector spans and is independent, so the complex dimension is one; the same set has different dimensions over different fields.
Worked example
Now allow complex scalars on the same set and recount the coordinates.
\[ V = \mathbb{C}, \quad \text{scalars from } \mathbb{C} \]
Reach every vector from a single one
Why: Any complex number is itself a complex scalar times the number one, so one vector already spans the whole set.
\[ z = z \cdot 1 \ \text{ for every } z \in \mathbb{C} \]
Confirm no smaller spanning set
Why: A single nonzero vector cannot be dropped without losing everything but the zero vector, so one vector is minimal.
\[ \{1\} \ \text{spans } \mathbb{C} \text{ and is independent} \]
Verify the count
Why: One vector spans and is independent, so the complex dimension is one; the same set has different dimensions over different fields.
\[ \dim_{\mathbb{C}} \mathbb{C} = 1 \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "C as a complex vector space", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: One vector spans and is independent, so the complex dimension is one; the same set has different dimensions over different fields.
Trap
Answering the dimension of C is two as if it were a fixed number attached to the set of complex numbers.
Dimension is stated without naming the field, so the answer is ambiguous and often wrong.
\[ \dim \mathbb{C} = 2 \ ? \]
Always name the scalar field. Dimension is a statement about the pair, the set and its field, never the set alone.
\[ \dim_{\mathbb{R}} \mathbb{C} = 2, \qquad \dim_{\mathbb{C}} \mathbb{C} = 1 \ \checkmark \]
Two truths and a lie
Sort into buckets
Some of these hold up and some are the exact mistakes this lesson is built to prevent. Sort them.
vector names a role, not a shape.; The scalars are not arbitrary. They form a field: you can add, subtract, multiply, and divide by anything nonzero.vector always means an arrow in two or three dimensions, so anything without a picture cannot be a vector.; Checking only that a set is closed under addition, then declaring it a vector space.Concept
The axioms never used division of scalars, only that scalars add and multiply. So the same list makes sense over any ring.
module — A set with addition and scalar multiplication by elements of a ring, satisfying the same eight axioms as a vector space. A vector space is exactly a module whose ring of scalars is a field.
Modules are the general theory; vector spaces are the tame special case where the scalars form a field.
Definition probe
Sort into buckets
Every line below is part of the definition of vector space or of module — one or the other, never both. Put each where it belongs.
Concept
Division is what makes bases exist. Over a field every vector space has a basis and a well-defined dimension.
Over a general ring this collapses. The integer pairs, viewed with integer scalars, form a module with no way to divide, and many familiar theorems simply fail.
\[ \mathbb{Z}^2 \ \text{is a } \mathbb{Z}\text{-module, not a vector space} \]
That is the deep reason this whole course fixes a field: division buys you coordinates.
Explain it
Discussion prompt
Explain Why fields are the sweet spot to a student a year behind you. No notation, no jargon they have not met — and it still has to be true.
Hint: If your explanation needs a symbol they have never seen, you are describing the notation rather than the idea.
Answer:
Division is what makes bases exist. Over a field every vector space has a basis and a well-defined dimension.
Concept
Non-examples usually break in one of three predictable places, and spotting which one is the whole skill.
The set may lack a zero vector, fail to be closed under addition, or fail to be closed under scalar multiplication. Test those three first before touching the associativity and distributive laws.
| failure | typical culprit |
|---|---|
| no zero vector | an affine or shifted set |
| not closed under + | degree exactly n; unit circle |
| not closed under scaling | integer lattice; first quadrant |
Comparison
Comparison matrix
From How a candidate fails: refill the typical culprit column from what you know. The rest of the table is as it appeared.
| failure | typical culprit |
|---|---|
| no zero vector | an affine or shifted set |
| not closed under + | degree exactly n; unit circle |
| not closed under scaling | integer lattice; first quadrant |
Step zero
Discussion prompt
Worked example: the first quadrant fails — before any calculation: what is the plan? Name the moves in order, in plain English, without doing the arithmetic.
Hint: It starts with: Check the zero vector
Answer:
Worked example
Take the pairs with both coordinates at least zero. Decide whether this is a real vector space.
\[ Q = \{ (x, y) : x \ge 0,\ y \ge 0 \} \]
Check the zero vector
Why: The origin has both coordinates zero, which satisfies both inequalities, so a zero vector is present and this is not where it breaks.
\[ (0,0) \in Q \]
Try scaling by a negative scalar
Why: Multiplying a nonzero point by negative one sends it into the third quadrant, where the coordinates are negative and the condition fails.
\[ (-1)\,(1,1) = (-1,-1) \notin Q \]
Verify the axiom that breaks
Why: The set is not closed under scalar multiplication and no point except zero has an additive inverse inside it, so it is not a vector space.
\[ Q \text{ is not closed under scaling} \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the first quadrant fails", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: The origin has both coordinates zero, which satisfies both inequalities, so a zero vector is present and this is not where it breaks.
Ranking
Put in order
Put the moves of Worked example: the cancellation law into the order they have to happen.
Why: These are the moves of the worked example in the order it makes them, and each one is set up by the one before it. Every vector has an additive inverse, so add the inverse of w to both sides of the equation.
Worked example
The axioms let you cancel a common vector from both sides of an equation, just as with numbers. Prove it holds in every vector space.
\[ \text{Claim: } u + w = v + w \ \Rightarrow\ u = v. \]
Add the negative of the common vector
Why: Every vector has an additive inverse, so add the inverse of w to both sides of the equation.
\[ (u + w) + (-w) = (v + w) + (-w) \]
Regroup with associativity
Why: Associativity lets w and its inverse meet; their sum is the zero vector, which then drops out.
\[ u + (w + (-w)) = v + (w + (-w)) \ \Rightarrow\ u + 0 = v + 0 \]
Verify with a concrete case
Why: Cancelling the common vector in the plane recovers the equal vectors, matching the general proof.
\[ (2,5)+(1,1)=(3,6)=(2,5)+(1,1) \ \Rightarrow\ (2,5)=(2,5) \ \checkmark \]
Picture it
Animation
Shows: Each line of the worked example "the cancellation law", appearing one at a time.
The same working the example does, in the order a tutor would write it.
Takeaway: Cancelling the common vector in the plane recovers the equal vectors, matching the general proof.
Concept
Three of our examples have no finite spanning set. They are the natural habitat of analysis and are not exotic.
All polynomials, all sequences, and all functions on an infinite set each need infinitely many independent directions.
\[ \mathcal{P}, \quad \mathbb{F}^{\mathbb{N}}, \quad \mathcal{F}(\mathbb{R}, \mathbb{R}) \]
The axioms treat them exactly like the plane. That uniform treatment is the payoff of working from axioms rather than pictures.
Analogy
Discussion prompt
Explain The infinite-dimensional spaces are everywhere by analogy to something with no Foundations of Higher Mathematics in it at all — a queue, a recipe, a map, a bank balance, whatever fits. Then say where your analogy breaks.
Hint: An analogy that never breaks is not an analogy, it is the same idea wearing a hat. Find the seam — that is the part that is actually new.
Answer:
Three of our examples have no finite spanning set. They are the natural habitat of analysis and are not exotic.
Intuition
In practice you almost never verify eight axioms from scratch. Most candidates sit inside a space you already trust.
When the candidate is a subset of a known vector space, the associativity, commutativity, and distributive laws are inherited automatically. Only three things can go wrong.
So the working test collapses to: is there a zero vector, and is the set closed under addition and under scaling. That is the recipe on the next slide.
Counterexample
Discussion prompt
In practice you almost never verify eight axioms from scratch. Most candidates sit inside a space you already trust.
That is stated as though it always holds. Do one of two things: produce a case where it fails, or say precisely what rules such a case out. "It just does" is not on the menu.
Hint: Hunt at the extremes first — zero, one, negative, empty, equal. If every extreme survives, the reason they survive is the proof.
Answer:
When the candidate is a subset of a known vector space, the associativity, commutativity, and distributive laws are inherited automatically. Only three things can go wrong.
Pattern
1. Identify the field and the two operations
Why: Nothing is a vector space in isolation; pin down the scalars and how addition and scaling are defined before testing anything.
2. Check that the zero vector is in the set
Why: A missing zero is the fastest disqualifier and immediately flags affine or shifted sets.
3. Check closure under addition
Why: Add two general members and confirm the result still satisfies the defining condition of the set.
4. Check closure under scalar multiplication
Why: Scale a general member by an arbitrary field element, including negatives and fractions, and confirm it stays inside.
5. If it sits inside a known space, you are done; otherwise verify the remaining axioms
Why: Subsets of a known vector space inherit the algebraic laws, so the three closure checks suffice; a from-scratch set needs the full eight.
Real world
Discussion prompt
Outside this lesson: where does Vector Spaces: Axioms & Examples actually turn up? Name one concrete situation — a job, a piece of software someone ships, a decision somebody has to make — and say which part of Pattern: is this a vector space? is doing the work in it.
Hint: Vague is the failure mode here. "Engineering" is not a situation; "deciding whether this build is fast enough to ship" is.
Answer:
That deck covers the eight vector-space axioms over a field, the consequences those axioms immediately force, and the zoo of genuine examples: tuples, matrices, polynomials, functions, the solution sets of homogeneous systems, and C regarded over different fields. It targets the misconceptions that a vector is an arrow in R^n, that any set of vectors is a space, that closure under scalar multiplication comes for free, and that the choice of scalar field does not affect dimension.
Elimination
Eliminate the wrong options
Which of these is NOT a real vector space?
3 of these 4 are wrong. Strike them one at a time, and say what rules each one out before you strike the next. The survivor is the answer.
Survives elimination: B
Why: Degree exactly three excludes the zero polynomial, so it has no zero vector, and two such polynomials can cancel leading terms and drop in degree, breaking closure under addition. The other three all contain zero and are closed.
Check
Each option uses the natural operations over the real numbers. Which one is not a vector space?
Check your understanding
Which of these is NOT a real vector space?
Answer: B
Why: Degree exactly three excludes the zero polynomial, so it has no zero vector, and two such polynomials can cancel leading terms and drop in degree, breaking closure under addition. The other three all contain zero and are closed.
Check
Let v be any vector in any vector space over a field, and let the scalar be the field's zero.
Check your understanding
What must the scalar zero times v equal?
Answer: A
Why: Writing the scalar zero as zero plus zero and applying the distributive law gives 0v equals 0v plus 0v; cancelling one copy forces 0v to be the zero vector in every vector space.
Check
Consider the complex numbers as a vector space, this time with the real numbers as the scalars.
Check your understanding
What is the dimension of C as a vector space over R?
Answer: B
Why: Every complex number is uniquely a real multiple of one plus a real multiple of the imaginary unit, and neither building block is a real multiple of the other, so two independent directions are needed over the reals.
Prediction
Predict first
Is the set of integer pairs a real vector space?
Answer it in your own words, now, with nothing to choose from. The options are on the next slide — and picking the right one off a list is an easier skill than producing it.
Correct: No, because it is not closed under scaling by real numbers
Why: Scaling an integer pair by a non-integer real, such as one half, produces a non-integer coordinate that leaves the set, so closure under scalar multiplication fails even though addition and the zero vector are fine.
Check
Take the pairs of integers inside the real plane, with the usual addition and with real scalars.
Check your understanding
Is the set of integer pairs a real vector space?
Answer: C
Why: Scaling an integer pair by a non-integer real, such as one half, produces a non-integer coordinate that leaves the set, so closure under scalar multiplication fails even though addition and the zero vector are fine.
Prediction
Predict first
Why is the line x plus y equals one not a vector space?
Answer it in your own words, now, with nothing to choose from. The options are on the next slide — and picking the right one off a list is an easier skill than producing it.
Correct: It has no zero vector and is not closed under addition
Why: The origin has coordinates summing to zero, not one, so the zero vector is absent, and adding two points on the line gives coordinates summing to two, which leaves the line, so closure under addition also fails.
Check
A line in the plane consists of all points whose coordinates sum to one. Which axiom does this set violate?
Check your understanding
Why is the line x plus y equals one not a vector space?
Answer: A
Why: The origin has coordinates summing to zero, not one, so the zero vector is absent, and adding two points on the line gives coordinates summing to two, which leaves the line, so closure under addition also fails.
Elimination
Eliminate the wrong options
Which three checks are sufficient for a subset of a known vector space?
3 of these 4 are wrong. Strike them one at a time, and say what rules each one out before you strike the next. The survivor is the answer.
Survives elimination: A
Why: A subset of a known vector space inherits the algebraic laws automatically, so only three things can go wrong: a missing zero vector, failure of closure under addition, or failure of closure under scalar multiplication.
Check
You want to verify a subset of the real plane is a vector space using the shortcut for subsets of a known space.
Check your understanding
Which three checks are sufficient for a subset of a known vector space?
Answer: A
Why: A subset of a known vector space inherits the algebraic laws automatically, so only three things can go wrong: a missing zero vector, failure of closure under addition, or failure of closure under scalar multiplication.
Connect it up
Draw it
One page, no notation unless you need it: draw how these connect — Pattern: is this a vector space? · A vector space is a set tied to a field · A vector is anything you can add and scale · The scalars come from a field · Two operations, both closed. Put an arrow wherever one of them is what makes another possible, and label the arrow with why.
Recap
A vector space is a set, a field of scalars, and two closed operations obeying eight axioms: four make addition an abelian group, and four tie scaling to the field.
The axioms force the free consequences you will use constantly: the zero vector is unique, the scalar zero annihilates every vector, scaling by negative one gives the additive inverse, and you can cancel a common vector.
The examples are a zoo unified by one definition: tuples, matrices, polynomials, functions, sequences, and solution sets of homogeneous systems all share every theorem.
To test a candidate, name the field, then check for a zero vector and closure under both operations. Missing zeros flag affine sets; the integer lattice and the first quadrant flag broken scaling.
Finally, the scalar field is part of the data: the complex numbers have real dimension two but complex dimension one. Next we study subspaces, sums, and direct sums inside a fixed space.
Want this taught 1-on-1? Alexander tutors Foundations of Higher Mathematics — $55/session, free consultation.