14
$\begingroup$

Do there exist functions $f : \mathbb{R} \rightarrow \mathbb{R}$ such that $f(x+y)=f(x)+f(y),$ but which aren't linear? I bet you they exist, but I can't think of any examples.

Furthermore, what hypotheses do we need to put on $f$ before no such functions exist? I feel continuity should be enough.

$\endgroup$
4
  • 2
    $\begingroup$ You might look at the Wikipedia article on the Cauchy functional equation. $\endgroup$ May 8, 2013 at 15:11
  • 1
    $\begingroup$ If you assume axiom of choice, these function is exists - you can construct this function to use Hamel basis of $\mathbb{R}$ over $\mathbb{Q}$. - But if you does not assume axiom of choice, you can't show that nonlinear additive function - If you assume axiom of determinacy, every real function is measurable. And every measurable additive real function is linear. $\endgroup$
    – Hanul Jeon
    May 8, 2013 at 15:18
  • $\begingroup$ This has got to be a dupe. $\endgroup$
    – Aryabhata
    May 8, 2013 at 17:01
  • 1
    $\begingroup$ relevant mathoverflow.net/questions/57426/… $\endgroup$
    – clark
    May 8, 2013 at 17:03

2 Answers 2

12
$\begingroup$

Yes continuity is enough: You can quickly show that $f(x)=x\cdot f(1)$ for $x\in\mathbb N$, then for $x\in\mathbb Z$ and then for $x\in\mathbb Q$; assuming continuity, this implies validity for all $x\in\mathbb R$.

Any other functions only exist per Axiom of Choice: View $\mathbb R$ as a vector space over $\mathbb Q$ and take any $\mathbb Q$-linear map (which need not be $\mathbb R$-linear).

$\endgroup$
1
  • $\begingroup$ how can you be sure about, some $\mathbb Q$ linear maps would not be $\mathbb R$ linear? $\endgroup$ Nov 14, 2023 at 8:02
2
$\begingroup$

Also, it is well known that graph $\{(x, f(x):x\in \Bbb R\}$ of every solution $f$ except $f(x)=ax$ is dense in the plane $\Bbb R^2$.

$\endgroup$

You must log in to answer this question.

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