(This is a question from an engineer and extremely naïve mathematician when it comes to the topic of irrational and transcendental numbers and the precise distinction between them.)

I begin with two examples of what I mean by the title question:

Let $A_1=(a+\sqrt{b})$ where a and b are rational numbers. Find $A_2$ so that the product $A_1 A_2$ is rational?

Obviously the answer is $A_2=(a-\sqrt{b})$ with $A_1 A_2=a^2-b$

Another slightly more difficult example with $a$, $b$ and $c$ rational is $A_1=(a+\sqrt{b}+\sqrt[3]{c}\,)$

with a little more work I found that one possible value for $A_2$ is

$$A_2=\left(\sqrt[3]{c} \left(-\left(a+\sqrt{b}\right)\right)+\left(a+\sqrt{b}\right)^2+c^{2/3}\right) \left(a^3-\sqrt{b} \left(3 a^2+b\right)+3 a b+c\right),$$ with $A_1 A_2=a^6-3 a^4 b+2 a^3 c+3 a^2 b^2+6 a b c-b^3+c^2$

Generalising further: Is it possible to prove that for every algebraic irrational number multiplier (real or complex) (or at least for a large subset of them), there exists at least one non-equal finite closed form multiplicand, so the product of the two is a rational number?

Implying that the inverse $\frac{1}{A_1}=\frac{A_2}{q}$ can be "sensibly" calculated in closed form (with $q$ rational). I am using the word "sensibly" to signify that I have no clear and precise mathematical definition for exact nature and productive constraints on this closed form, just the examples above.

The underlying reason I ask is that it is not immediately clear to me that an irrational (possibly transcendental?) number like $(\zeta(3)+a)$ or $(\zeta(3)+\sqrt{a})$, where $a$ is rational, has an associated non-equal closed form multiplicand, with the product of the two, resulting in a rational number.

  • $\begingroup$ What do you mean for non-equal, finite, closed form multiplicand? Why can't you just say $A \frac{1}{A} = 1$? $\endgroup$ – G Aker Apr 14 at 21:19
  • $\begingroup$ @GAker: It seems to me that for $(\zeta(3)+a) \frac{1}{\zeta(3)+a}=1$, for example, I can simply define a new closed form $\frac{1}{\zeta(3)+a}$ in some clever inventive way. It is not then clear to me how a matched pair of such numbers are then proved consistent with a wider algebra involving other existing forms of irrational algebraic numbers. $\endgroup$ – James Arathoon Apr 14 at 21:59
  • $\begingroup$ Again I'm not sure what you mean by an algebra in this context. I'll have a go answering what I think you're asking. If a number is of the form $a = a_1x_1+a_2x_2 + a_3x_3+...+ a_nx_n$ where the $x_i$ are the basis of an algebraic extension of $\mathbb{Q}$ and the $a_i$ are all rational numbers then $1/a$ can be written in the same form. This is because extension fields are still fields. Think I might be misunderstanding the question however. $\endgroup$ – G Aker Apr 15 at 7:22

For each non-zera irrational algebraic number $\alpha$, $\dfrac1\alpha$ is also an algebraic irrational number and $\alpha\times\dfrac1\alpha=1\in\mathbb Q$.


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.