Let $S$ be a subspace of a normed linear space, $X$ and $x_0\in X\backslash S$. Consider the subspace spanned by $M,$ i.e. \begin{align} M:=[S\cup \{x_0\}]=\{m=x+\alpha\,x_0:\,x\in S,\;\text{for some}\;\alpha\in\Bbb{R}\} \end{align} I want to show that $m$ is unique.


Let $m\in M,$ then there exists $\alpha\in\Bbb{R} $ such that $m=x+\alpha\,x_0.$ Suppose we have another representation, then there exists $\beta\in \Bbb{R},\;\beta\neq \alpha,$ such that $m=x+\beta\,x_0.$ Thus, \begin{align} x_0=0 \;\text{which implies}\;x_0\in S,\;\text{contradiction, since }\; x_0\notin S.\end{align} Please, I'm I right? If not, could you please, provide an alternative proof?

  • $\begingroup$ I'm not 100% sure I understand the question. As I understand it, you're trying to show that, if you have $x + \alpha x_0 = y + \beta x_0$, where $x, y \in S$, then $x = y$ and $\alpha = \beta$? $\endgroup$ – Theo Bendit Jan 10 at 5:01
  • $\begingroup$ @Theo Bendit: That's true! That almost shows that I'm not correct! $\endgroup$ – Omojola Micheal Jan 10 at 5:02
  • $\begingroup$ You're almost correct; instead consider $x - y = (\beta - \alpha)x_0$. $\endgroup$ – Theo Bendit Jan 10 at 5:03
  • $\begingroup$ @Theo Bendit: Okay, let me try that! $\endgroup$ – Omojola Micheal Jan 10 at 5:04
  • $\begingroup$ If you get, write an answer. :-) $\endgroup$ – Theo Bendit Jan 10 at 5:05

Credits to @Theo Bendit for the hints.


Let $x_0\in X\backslash S$ be arbitrary. Suppose that $m\in M, $ then there exists $x\in S\;\text{and}\;\alpha\in\Bbb{R} $ such that $m=x+\alpha\,x_0.$ Assume there is another representation, then there exists $y\in S\;\text{and}\;\beta\in \Bbb{R}$ such that $m=y+\beta\,x_0,$ where $x\neq y$ or $\alpha\neq\beta$. If $x\neq y$, then $\beta \neq \alpha.$ Otherwise, if $\beta \neq \alpha$ then $x=y$ or $x\neq y.$ In both cases, $\beta \neq \alpha$. Thus, \begin{align} \left(x+\alpha\,x_0=y+\beta\,x_0\right)&\implies (x-y)=(\beta-\alpha)\,x_0\\&\implies x_0=\dfrac{1}{\beta-\alpha}(x-y)\in S,\;\text{where}\;\beta-\alpha\neq 0\;\text{and}\;x-y\in S,\\&\implies x_0\in S,\;\text{since }\;S\;\text{is a subspace of a normed linear space }\\&\implies\text{a contradiction }\end{align} Hence, the representation $m=x+\alpha x_0,\;$ for $\;m\in M$ is unique.

  • 1
    $\begingroup$ You should make the final implication more clear. What's wrong with $\alpha - \beta \neq 0$? You should be able to use this assumption to conclude $x_0 \in S$ (though you won't be able to conclude $x_0 = 0$). $\endgroup$ – Theo Bendit Jan 10 at 5:17
  • $\begingroup$ @Theo Bendit: Okay, thanks for the guidiance. $\endgroup$ – Omojola Micheal Jan 10 at 5:18
  • $\begingroup$ Hint: subspaces are closed under subtraction and scalar multipliaction. $\endgroup$ – Theo Bendit Jan 10 at 5:38
  • $\begingroup$ @Theo Bendit: Okay, thanks, once again, for the guidiance. You have my respect! $\endgroup$ – Omojola Micheal Jan 10 at 5:51
  • 1
    $\begingroup$ @Theo Bendit: Thanks a lot for your time and efforts. $\endgroup$ – Omojola Micheal Jan 10 at 6:11

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.