The octonions are a noncommutative nonassociative normed division algebra over $\mathbb{R}$. Multiplication distributes over addition. Somehow, the existence of a norm implies the existence of multiplicative inverses.

Since multiplication is not associative, the nonzero octonions are not a group, but a loop. I wonder if the octonions can still be regarded as a "field", though, or if there's a more proper name for them.

  • 5
    $\begingroup$ A "non-associative division algebra" maybe? $\endgroup$ Commented Apr 27, 2017 at 4:52
  • 2
    $\begingroup$ Fields are basically always defined to be associative and almost always defined to be commutative. But, paradoxically, Lord Shark the Unknown's comment is true that it is less rare for "division algebras" to omit associativity from the list of axioms. $\endgroup$
    – rschwieb
    Commented Apr 27, 2017 at 14:56

1 Answer 1


You have answered your own question since a field must have an associative product, yet the octonions are not associative.

Although multiplication is not associative, it is alternative and flexible.

Suppose $x$ and $y$ are elements of the octionions $\mathbb{O}$.


  1. $(xx)y=x(xy)$
  2. $x(yy)=(xy)y$
  3. $(xy)x=x(yx)$

The first two properties make $\mathbb{O}$ alternative the third makes it flexible.

  • $\begingroup$ The usual definition of division ring also requires associativity. It's the Quaternions that are notable for being a division ring. $\endgroup$
    – user14972
    Commented Apr 27, 2017 at 6:22
  • $\begingroup$ @Hurkyl, You are correct, I will amend my answer. $\endgroup$ Commented Apr 27, 2017 at 14:49

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .