Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Suppose we have a complex vector space $V$.

A norm is a function $f : V \rightarrow \mathbb{R}$ which satisfies

(i) $f(x) \ge 0$ for all $x \in V$ (Positivity - Non-Negativity)

(ii) $f(x + y) \le f(x) + f(y)$ for all $x, y \in V$ (Subadditivity - Triangle Inequality)

(iii) $f(\lambda x) = |\lambda|f(x)$ for all $\lambda \in \mathbb{C}$ and $x \in V$ (Positive Homogeneity)

(iv) $f(x) = 0$ if and only if $x = 0$ (Definiteness)

From these properties, we define the $l_p-norm$

$$ ||x||_p = \left(\sum \limits_{i=1}^{n}|x_i|^p \right)^\frac{1}{p}$$

My questions relate to why these four properties were chosen for this definition:

(1) If norm is just a measure, why do we need all four of these properties?

(2) Do these properties allow for a nice geometric interpretation of norms?

(3) Were these properties chosen to nicely allow for a nesting of the various norms? For example

$$ ||x||_\infty \le ||x||_2 \le ||x||_1$$

(4) Were these properties chosen to satisfy pairwise inequalities? For example $$ ||x||_1 \le n||x||_\infty$$

(5) Were these properties chosen because they lend themselves in other variants? For example, in $l_2$ for bi-infinite sequences, vector-valued sequences or general sequences.

(6) Do these properties allow an extension into other function spaces where all of the items above still hold? For example, $L_2$ or $L_p$ function spaces.

I haven't looked at norms in quite a while, so please forgive me if I am missing something very obvious.

share|cite|improve this question
I'll have to find my link to a math etymology site, which addresses first uses and subsequent uses of terms, when they were introduced, by whom, to represent what, etc. ... If I remember, I'll post a link here to it. But yes, I always wonder about how/when/why "conventions", terms, notation, ideas, first came into play, and how they've morphed over time. – amWhy Apr 19 '13 at 2:33
@amWhy: Those are some excellent questions and it would be nice to have something that captures all of that, as well as a detailed list of notation and usage over time. – Amzoti Apr 19 '13 at 2:34
up vote 8 down vote accepted

The intuition behind the definition of norm is that the norm of a vector is a way to measure the length of the vector. Then the definition becomes clear:

  • The length should be non-negative.
  • In a triangle, the length of one side is smaller than the sum of the lengths of the other sides (this where the name triangle inequality comes from).
  • If we stretch a vector, i.e. if we multiply it by a number $\lambda$, the length should be multiplied by $|\lambda|$.
  • The only vector with length $0$ should be the zero vector.

Moreover, in a normed space you can define a distance as $$ d(x,y)=\|x-y\|. $$ This gives the space a structure of metric (and hence topological) space, allowing the definition of open and closed sets, limits,...

There are many normed spaces used throughout mathematics:

  • $\mathbb{R}^n$ and $\mathbb{C}^n$.
  • The space of continuous functions $f\colon K\to\mathbb{C}$ where $K$ is a compact topological space with the uniform norm.
  • The $L^p$ and $\ell^p$ spaces, $1\le p\le\infty$.
  • Sobolev spaces.
  • ...
share|cite|improve this answer

A possible motivation for three of the four properties begins with the idea that $\Vert x\Vert$ should be the distance between the points $0$ and $x$, or equivalently, the length of the vector $x$. Now the vector from the point $x$ to the point $y$ is $y-x$, so $\Vert y-x\Vert$ should be the distance between those two points. That means the function $x,y\mapsto\Vert y-x\Vert$ should be a metric on the vector space. In order for it to satisfy the axioms of a metric, you need properties 1, 2, and 4 on your list.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.