That is, how to prove the following identity: $$a \times (b+c) = a \times b + a \times c$$ where the $\times$ represents cross product of two vectors in 3-dimensional Euclidean space.


To see it geometrically:

Recall that $\lVert x\times y\rVert$ represents the area of the parallelogram with sides $x,y$. If you then "glue" together the parallelograms with sides $a,b$ and $a,c$ along the side $a$, you get a hexagon (not regular) with the same area of the parallelogram with sides $a,b+c$.

Sum of parallelograms.

Edit: As another user correctly noted, the above proof works only when $a$ lies in the plane spanned by $b$ and $c$. If not, then we can reduce to this case by considering the triangle $T$ with basis $b+c$ and sides $b,c$, as suggested by this image1:

sum of bivectors

Indeed, without loss of generality we may assume that $a$ is orthogonal to $b$ and $c$. Then the angle and proportion between $a \times b$ and $a \times c$ are the same as those between $b$ and $c$, because in this case the cross product with $a$ is equivalent to the composition of a rotation by $90$ degrees and a dilation by $\lVert a \rVert$. Therefore if $P$ is the prism with base $T$ and height $a$, then the area of the projection of the face $[a,b]$ on the face $[a,b+c]$ is the same as the length of the cross product between $a$ and the projection of $b$ on $b+c$, and similarly for the face $[a,c]$.

1. Courtesy of WikiMedia.

  • $\begingroup$ I added a descriptive image. $\endgroup$ – A.P. Apr 15 '13 at 17:12
  • $\begingroup$ Could whoever downvoted this explain why they did so, allowing me to improve my answer? $\endgroup$ – A.P. Apr 5 '15 at 12:26
  • 1
    $\begingroup$ Your answer assumes that $a$, $b$ and $c$ all lie in the same plane. When they don't, which they don't in general, it breaks down. $\endgroup$ – Christopher Apr 7 '15 at 12:51
  • $\begingroup$ ($a \times b$ and $a \times c$ are no longer parallel, so the magnitude of their sum is less than the sum of their magnitudes; the picture at the bottom fails because it's now depicting a 3-d solid rather than a planar shape.) $\endgroup$ – Christopher Apr 7 '15 at 12:53
  • $\begingroup$ @user73985 thanks for catching the omission. I amended my answer and it should be fine now. $\endgroup$ – A.P. Apr 10 '15 at 19:28

Just to enrich the post for future readers, I would like to add another derivation that I found on the internet. This proof uses the distributivity of the dot product (which is easier to prove), and the property that the circular commutation of vectors doesn't change the triple product of the vectors (which is quite obvious, since the triple product is just the volume of the parallelepiped formed by the vectors).

Let $d = a \times (b + c) - a \times b - a \times c$

so it is required to prove that $d = 0$:

$d^2$ = $d \cdot d$

$= d \cdot (a \times (b + c) - a \times b - a \times c)$

$= d \cdot (a \times (b + c)) - d \cdot (a \times b) - d \cdot (a \times c)$

$= (d \times a) \cdot (b + c) - (d \times a) \cdot b - (d \times a) \cdot c$

$= (d \times a) \cdot (b + c) - (d \times a) \cdot (b + c)$

$= 0$

Therefore $d = 0$, so $a \times (b + c) = a \times b + a \times c$.

  • 2
    $\begingroup$ I think you need to add a bit more work for this proof to be convincing. Namely, prove that (i) the triple product is the volume of the parallelepiped; and (ii) the triple product is circularly commutative without relying on the distributivity of the cross product. Otherwise one wonders if these latter two facts somehow depend on the fact that the cross product is distributive. (I know they don't, but there is this gap in your answer.) $\endgroup$ – Kenny LJ Apr 22 '16 at 5:15

Let $a=(a_1,a_2,a_3),~~b=(b_1,b_2,b_3),~~c=(c_1,c_2,c_3)\in\mathbb R^3$ so $$a\times(b+c)=\begin{vmatrix} i\;\;\;j\;\;\;k \\ a_1\;\;\;a_2\;\;\;a_3 \\ b_1+c_1\;\;\;b_2+c_2\;\;\;b_3+c_3 \end{vmatrix}=i\left(a_2b_3+a_2c_3-a_3b_2-a_3c_2\right)-j(...)+k(...)$$ Now try to rearrange the above terms to find the result. See that in the first term we have $i\left(a_2b_3+a_2c_3-a_3b_2-a_3c_2\right)=i(a_2b_3-a_3b_2)+i(a_2c_3-a_3c_2)$.

  • 6
    $\begingroup$ I think, in this proof, the conclusion is implicitly assumed. $\endgroup$ – Berci Apr 15 '13 at 9:25
  • $\begingroup$ @Berci: As you commented before, I don't know what the OP know and how he/she wanted to do the problem. $\endgroup$ – mrs Apr 15 '13 at 9:27
  • 1
    $\begingroup$ Doesn't this expression (as a determinant) for the cross product of two vectors itself require distributivity of the cross product? $\endgroup$ – SamInuyasha ANMF Apr 22 '18 at 10:47

Found a geometric proof in https://en.wikiversity.org/wiki/Cross_product. Refer to section: Equivalence of the two Definitions

  • 2
    $\begingroup$ The question has already an accepted answer. I'm not sure your answer is highly useful, in my opinion. $\endgroup$ – Watson Feb 4 '16 at 13:29
  • $\begingroup$ This. The proof is straightforward if you project a and b onto the plane P orthogonal to c. $\endgroup$ – PJ_Finnegan Oct 24 at 19:17
  • $\begingroup$ This should be the accepted answer. $\endgroup$ – PJ_Finnegan Oct 24 at 19:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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