9
$\begingroup$

In our algebra class, our teacher often does the following:

$a + b\sqrt{2} = 5 + 3\sqrt{2} \implies \;\text{(by inspection)}\; a=5, b = 3 $

I asked her why we can make this statement. She was unable to provide a satisfactory answer. So I tried proving it myself.

$a + b\sqrt{2} = x + y\sqrt{2}$. We are required to prove that $a = x$, and $b = y$. Manipulating the equation, we get $\sqrt{2}(b - y) = x - a$, or $\sqrt{2} = \frac{x-a}{b-y}$. Expanding this, we get $\sqrt{2} = \frac{x}{b-y} + \frac{a}{b-y}$. I tried various other transformations, but nothing seemed to yield a result.

$\endgroup$
1
  • 4
    $\begingroup$ (+1) for not accepting the "inspection" claim of your teacher without further arguments! $\endgroup$ – azimut Aug 9 '13 at 11:59
3
$\begingroup$

We want to show that the only rational solutions $a,b,x,y$ of $$a + b\sqrt{2} = x + y\sqrt{2}$$ are given by $a = x$ and $b = y$.

(Note that if you allow for real values of $a,b,x,y$, then for example $a = x + \sqrt{2}$, $b = y - 1$ would be a solution, too.)

If $y = b$, then obviously $a = x$.

Now assume $y \neq b$. Then $$a + b\sqrt{2} = x + y\sqrt{2}\\ \implies a-x = (y-b)\sqrt{2}\\ \implies (a-x)^2 = 2(y-b)^2\\ \overset{y - b \neq 0}{\implies} 2 = \left(\frac{a-x}{y-b}\right)^{\!2}.$$ So $2$ is the square of a rational number. This contradicts the fact that $\sqrt{2}$ is irrational.


By the above argument, we have shown that the only rational solution of $a + b\sqrt{2} = 0$ is $a = b = 0$. In terms of linear algebra, this property is formulated as "$1$ and $\sqrt{2}$ are $\mathbb Q$-linearly independent". From this it follows that any representation $a + b\sqrt{2}$ with $a,b\in\mathbb Q$ uniquely determines $a$ and $b$.

$\endgroup$
6
  • $\begingroup$ But, what if $a = x, b = y$? The equation would become, $2 = \left(\frac{a-a}{b-b}\right)^2$, which makes no sense at all. $\endgroup$ – Gerard Aug 9 '13 at 12:00
  • 1
    $\begingroup$ @AnanayAgarwal: The case $b = y$ is treated separately. In this place I'm assuming that $b\neq y$. $\endgroup$ – azimut Aug 9 '13 at 12:01
  • $\begingroup$ Why do you know $(a-x)/(y-b)$ is rational? Are you make that assumption on the coefficients? $\endgroup$ – Clayton Aug 9 '13 at 12:02
  • $\begingroup$ You are prooving there's only one solution made up of rationals, but what if $x,y$ are irrational? $\endgroup$ – MyUserIsThis Aug 9 '13 at 12:04
  • $\begingroup$ @clayton I assume that $a,b,x,y$ are rational. Otherwise, you get more solutions (I added a comment on this), so I guess this assumption was also made by the OP. $\endgroup$ – azimut Aug 9 '13 at 12:06
3
$\begingroup$

HINT:

Assuming $a,b,x,y$ are rationals $\sqrt2(b-y)=x-a$ rational which is only possible if $b-y=0$

As for $b-y\ne0,\sqrt 2=\frac{x-a}{y-b}$ which is rational

$\endgroup$
2
  • 2
    $\begingroup$ Oh! I get it now. My teacher didn't tell me that this could be only possible if the numbers in question were rationals. That's why I thought there was a certain ambiguity to the method. $\endgroup$ – Gerard Aug 9 '13 at 11:58
  • 1
    $\begingroup$ @AnanayAgarwal, please find the edited answer and related math.stackexchange.com/questions/457382/… $\endgroup$ – lab bhattacharjee Aug 9 '13 at 12:02

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.