1
$\begingroup$

Which of the following numbers is constructible?

1) $3.14141414\ldots$

2) $\sqrt{3}$

3) $5^\frac{1}{4}$

4) $2^\frac{1}{6}$

Also,

Given a segment of length $\pi$, is it possible to construct, with a straight edge and compass, a segment of length $1$?

I dont need full on proofs, just a little explanationif they could be constructed or not.

i have a feeling number 2 and 4 are constructible, just a educated guess because we could get $\sqrt{2}$ from a unit square and the diagonal is that and then we can just extend it I believe.

I just dont get it. Please help out

$\endgroup$
  • 1
    $\begingroup$ $3.141414\ldots$ is not irrational. It is equal to $$\frac{311}{99}$$ $\endgroup$ – apnorton Mar 7 '13 at 23:27
  • $\begingroup$ I did not know that. Thank you sir $\endgroup$ – MathGeek Mar 7 '13 at 23:28
5
$\begingroup$

I'll only answer the last part, because everything else was done by anorton. Given a segment of length $1$ you can't construct a segment of length $\pi$ -- this is what we mean when we say $\pi$ is not constructable.

Starting with a segment of length $\pi$ is just a re-scaling of the constructability problem by a factor of $\pi$. The constructable numbers starting from $\pi$ are precisely the constructable numbers starting with $1$ multiplied by a factor of $\pi$. You can only construct $1$ from $\pi$ if you can construct $1/\pi$ from $1$. This, however, is impossible -- constructable numbers form a field, so if you could construct $1/\pi$ you could also construct $\pi$.

$\endgroup$
3
$\begingroup$

One is most certainly constructable, as all rationals are constructable, and $3.1414\ldots = \frac{311}{99}$.

Two is also constructable, as we can construct an equilateral triangle of side length 2, and its height is $\sqrt{3}$.

We can construct $\sqrt{5}$, so we should be able to construct $\sqrt{\sqrt{5}}=5^\frac{1}{4}$.

I am not sure about $2^\frac{1}{6}$, but I do know $2^\frac{1}{3}$ is not constructable, so I doubt the former is. EDIT: as André Nicolas points out in the comment below, it is not.

As to the $\pi$ question, I don't know.

$\endgroup$
  • 1
    $\begingroup$ If $2^{1/6}$ were constructible, its square $2^{1//3}$ certainly would be, since we can multiply by straightedge and compass. $\endgroup$ – André Nicolas Mar 7 '13 at 23:41
  • $\begingroup$ for number three, how could we construct $\sqrt{5}$ and how does that imply $\sqrt{\sqrt{5}}$ confused about that actually. $\endgroup$ – MathGeek Mar 8 '13 at 0:05
  • $\begingroup$ @MathGeek Take a look at this question: math.stackexchange.com/questions/705/… $\endgroup$ – apnorton Mar 8 '13 at 0:09
3
$\begingroup$

This answer uses the algebraic characterization of constructible numbers, as in this Wikipedia page.

The minimal polynomial of $2^\frac{1}{6}$ is $x^6-2$ (which is irreducible by Eisenstein's criterion), so $\mathbf{Q}(2^\frac{1}{6})$ is a degree 6 extension. This means there is no tower whose intermediate extensions are all degree 2, so $2^\frac{1}{6}$ is not constructible.

The other ones have towers of quadratic extensions:

  • $\mathbf{Q}(5^\frac{1}{4})\supset \mathbf{Q}(5^\frac{1}{2})\supset\mathbf{Q}$ is a tower where every extension is degree 2, so $5^\frac{1}{4}$ is constructible.
  • $\mathbf{Q}(\sqrt{3})\supset \mathbf{Q}$ is degree 2, so $\sqrt{3}$ is constructible.
  • 3.1414141414... is constructible because it is rational.

Finally, if we could construct 1 from a segment of length $\pi$, then we could construct a segment of length $\frac{1}{\pi}$ from a segment of length 1; but $\frac{1}{\pi}$ is transcendental so this is impossible.

$\endgroup$

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.