1
$\begingroup$

I read that it is only by convention that $\sqrt{}$ means “the positive square root of”. The downvotes and user 'Thomas' 's answer compel me to clarify that I asked this hypothetical question only because of curiosity, not because of any desire to desecrate $\sqrt{}$.

What if $\sqrt{}$ meant “the NEGATIVE square root of”? Why might convention not have caused this?


I already understand the following and already read: 2013/10/23, 2013/11/16, 2014/5/26, 2015/5/12, 2015/9/24.

Source: Page A7, Appendix A, Calculus Early Transcendentals (6th ed.; 2008) by James Stewart:

Recall that the symbol $\sqrt{}$ means “the positive square root of.”
Thus $\sqrt{r} = s$ means $s^2 = r$ and $s \ge 0$.
Therefore, $\color{darkred} { \text { the equation $\sqrt{a^2} = a$ is not always true. It is true only when $a \ge 0$ } } $.
If $a < 0$, then $ -a > 0$, so we have $\sqrt{a^2} = -a$.
[...] we then have [...] $\sqrt{a^2} = |a|$.

$\endgroup$
  • 13
    $\begingroup$ It would just be an annoying change in notation. I'd have to start writing $- \sqrt{a}$ instead of $\sqrt{a}$ for no reason. $\endgroup$ – Qiaochu Yuan Oct 30 '15 at 20:30
  • 8
    $\begingroup$ There is already an operator that does this. It's denoted $-\sqrt{}$. $\endgroup$ – David Hill Oct 30 '15 at 20:31
  • 7
    $\begingroup$ We would no longer have $\sqrt{xy}=\sqrt x\sqrt y$, for example (take $x=y=2$). This seems like a desirable property to me. $\endgroup$ – Clayton Oct 30 '15 at 20:31
  • $\begingroup$ This is just one of the 2 branches of $\exp{\frac{1}{2} Log z}$. It is only a convinient way to denote The positive number $a$ that satisfies $a^2 = x$. $\endgroup$ – Ranc Oct 30 '15 at 20:40
3
$\begingroup$

You are free to change notation. The only thing that is going to happen is that it will confuse and annoy a lot of people. Also, if you use notation that is contrary to established notation without giving a good reason, people aren't going to take your serious as a mathematician.

As mentioned in the comments above, if you define $\sqrt{a}$ to be the negative number $b$ such that $b^2 = a$, then we would just have to put a minus in front of all square roots appearing in the literature. It would (as also mentioned in the comments above) cause the problem that $$ \sqrt{ab} \neq \sqrt{a}\sqrt{b} $$ And then you have have to make sense of stuff like $x^{1/3}$. How is this now defined? There are good reasons for picking the notation we have.

So basically, you have to ask yourself: why would you want to do that?

$\endgroup$
  • $\begingroup$ +1. Thank you, but your first paragraph sounds critical of my question, which I just elucidated. Does my emendment help? $\endgroup$ – Greek - Area 51 Proposal Oct 30 '15 at 20:48
  • $\begingroup$ @LePressentiment: Sorry if I sounded critical of your question. I decided to answer it because it does make sense to ask. The "What if" part of the question is basically answered by: nothing would happen except you would need to replace all $\sqrt{}$ by $-\sqrt{}$ in expressions. You would also not have various formulas true anymore. $\endgroup$ – Thomas Oct 30 '15 at 23:17
  • $\begingroup$ No problem; thank you for your elucidation! I just wished to confirm my respect for $\sqrt{}$. $\endgroup$ – Greek - Area 51 Proposal Nov 1 '15 at 0:35
3
$\begingroup$

So imagine that convention caused $\sqrt{}$ to mean “the NEGATIVE square root of”. Then what would change or differ?

We would get negative signs everywhere instead of positive ones. This will happen every-time we solve an equation with even powers.

Would mathematics be fundamentally changed?

Good question, but I don't think it would, only the signs will change.

$\endgroup$
2
$\begingroup$

As commenters said, mathematics would be the same up to the substitution $\sqrt{\ } \mapsto -\sqrt{\ }$.

Actually, a similar phenomenon already happens in the context of differential geometry and PDEs. The same symbol $\Delta$ means $-\frac{\partial^2}{\partial x_1^2} -\ldots -\frac{\partial^2}{\partial x_n^2}$ in the geometers's literature and $\frac{\partial^2}{\partial x_1^2} +\ldots +\frac{\partial^2}{\partial x_n^2}$ in the PDEs literature. It's a bit confusing but people are used to it.

$\endgroup$

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.