7

$$ \mathbb{C} \otimes_{\mathbb{R}} \mathbb{R} = \;? $$

I guess this guy is just $\mathbb{C}$, is this correct?

Abramo
  • 6,917

4 Answers4

6

Yes this is correct. One way of seeing this is considering the map $$\mathbb C\otimes_\mathbb R \mathbb R\to \mathbb C$$ which maps $$z\otimes x\mapsto xz$$ and showing that it is an isomorphism. Another way is to see that $\mathbb C\otimes_\mathbb R \mathbb R$ is $2$-dimensional as a real vector space and that the induced product is the same as the product of $\mathbb C$. Hence they are isomorphic as algebras.

Simon Markett
  • 10,636
  • 1
  • 22
  • 32
5

Yes. In fact, for any commutative ring $R$ with module $M$, you can say this about the tensor product with free modules: $$M\otimes_R(\oplus_{i=1}^n R)\cong\oplus_{i=1}^n M$$

rschwieb
  • 153,510
  • 1
    funny we have the exact same reputation...I wonder whether this question changes this ;) – Simon Markett May 25 '12 at 10:42
  • 1
    @SimonMarkett Haha, quite a coincidence! Now I see my speculation that these icons are generated directly from rep is incorrect (I really should ask around and look for the true source.) – rschwieb May 25 '12 at 10:48
  • 2
    Now guys I don't know which of your answer I should accept. I don't want to ruin this symmetry! ;-) – Abramo May 25 '12 at 10:54
  • @rschwieb If you mean the badges by "icons" then you will find an explanation under "badges" ;) – Simon Markett May 25 '12 at 10:58
  • @SimonMarkett No, I was referring to the colorful fractalish looking square picture which is everchanging. – rschwieb May 25 '12 at 10:59
  • @rschwieb: AFAIK they come from gravatar and are math.SE just happens to use gravatar, your reputation couldn't influence your avatar. – Najib Idrissi May 25 '12 at 11:27
2

Recall that given commutative rings $S$, $R$ together with a ring homomorphism $f:R \longrightarrow S$ and an $R$ - module $M$ we can consider the tensor product

$$S \otimes_R M.$$

Now $S$ can be considered as an $R$ - module by defining the action of $R$ on $S$ by $r\cdot s = f(r)s$ so that $S \otimes_R M$ has the structure of an $R$ - module. However what is even more interesting is that there is a unique $S$ - module structure on $S \otimes_R M$; we may define $S$ - multiplication by

$$s_1( s_2 \otimes m ) = (s_1s_2) \otimes m$$

for all $s_1,s_2 \in S$ and $m \in M$. One still needs to check that this is well defined and extends from elementary tensors to all tensors. This process of turning an $R$ - module like that into an $S$ - module is also known as extension of scalars.

Now back to your problem. We already know that $\Bbb{C} \otimes_\mathbb{R} \Bbb{R}$ has the structure of an $\Bbb{R}$ - module. By extension of scalars, we can consider this as a $\Bbb{C}$ - module. Now $\Bbb{R}$ as a module over itself is a free $\Bbb{R}$ - module with basis just the real number $1$ say. Then it is possible to show that the $\Bbb{C}$ - module

$$\Bbb{C} \otimes_\Bbb{R} \Bbb{R} $$

has basis $1 \otimes 1$. However a $\Bbb{C}$ - module is nothing more than a complex vector space and in our case a vector space that is of dimension 1 (over $\Bbb{C}$). Because two vector spaces over the same field of equal dimension are isomorphic we conclude that

$$\Bbb{C} \otimes_\Bbb{R} \Bbb{R} \cong \Bbb{C}$$

as $\Bbb{C}$ - modules.

  • Do we need extension of scalars here? Isn't $M = \mathbb C$ an $R = \mathbb R$-module? So we could use the (natural) isomorphism $M \otimes R \cong M$ directly. – Rudy the Reindeer Jun 15 '12 at 21:47
  • @MattN. I suppose you could. However I just took a different approach by considering everything as complex vector spaces from the beginning. –  Jun 15 '12 at 21:51
2

You could use that you have a (natural) isomorphism $M \cong M \otimes_R R$, see here for a proof.

Then $M = \mathbb C$ is an $R$-module for $R = \mathbb R$ and you get (as you suspected) that $\mathbb C \cong \mathbb C \otimes_{\mathbb R} \mathbb R$.